diff options
Diffstat (limited to 'utils/tuning')
23 files changed, 1239 insertions, 104 deletions
diff --git a/utils/tuning/README.rst b/utils/tuning/README.rst index ef3e6ad7..89a1d61e 100644 --- a/utils/tuning/README.rst +++ b/utils/tuning/README.rst @@ -1,11 +1,20 @@ .. SPDX-License-Identifier: CC-BY-SA-4.0 -.. TODO: Write an overview of libtuning +libcamera tuning tools +====================== -Dependencies ------------- +.. Note:: The tuning tools are still very much work in progress. If in doubt, + please ask on the mailing list. + +.. todo:: + Write documentation + +Installation of dependencies +---------------------------- + +:: + # Using a venv + python3 -m venv venv + . ./venv/bin/activate + pip3 install -r requirements.txt -- numpy -- opencv-python -- py3exiv2 -- rawpy diff --git a/utils/tuning/config-example.yaml b/utils/tuning/config-example.yaml new file mode 100644 index 00000000..1b7f52cd --- /dev/null +++ b/utils/tuning/config-example.yaml @@ -0,0 +1,12 @@ +general: + disable: [] + plot: [] + alsc: + do_alsc_colour: 1 + luminance_strength: 0.5 + awb: + greyworld: 0 + macbeth: + small: 1 + show: 0 +# blacklevel: 32
\ No newline at end of file diff --git a/utils/tuning/libtuning/ctt_awb.py b/utils/tuning/libtuning/ctt_awb.py new file mode 100644 index 00000000..abf22321 --- /dev/null +++ b/utils/tuning/libtuning/ctt_awb.py @@ -0,0 +1,378 @@ +# SPDX-License-Identifier: BSD-2-Clause +# +# Copyright (C) 2019, Raspberry Pi Ltd +# +# camera tuning tool for AWB + +import matplotlib.pyplot as plt +from bisect import bisect_left +from scipy.optimize import fmin +import numpy as np + +from .image import Image + + +""" +obtain piecewise linear approximation for colour curve +""" +def awb(Cam, cal_cr_list, cal_cb_list, plot): + imgs = Cam.imgs + """ + condense alsc calibration tables into one dictionary + """ + if cal_cr_list is None: + colour_cals = None + else: + colour_cals = {} + for cr, cb in zip(cal_cr_list, cal_cb_list): + cr_tab = cr['table'] + cb_tab = cb['table'] + """ + normalise tables so min value is 1 + """ + cr_tab = cr_tab/np.min(cr_tab) + cb_tab = cb_tab/np.min(cb_tab) + colour_cals[cr['ct']] = [cr_tab, cb_tab] + """ + obtain data from greyscale macbeth patches + """ + rb_raw = [] + rbs_hat = [] + for Img in imgs: + Cam.log += '\nProcessing '+Img.name + """ + get greyscale patches with alsc applied if alsc enabled. + Note: if alsc is disabled then colour_cals will be set to None and the + function will just return the greyscale patches + """ + r_patchs, b_patchs, g_patchs = get_alsc_patches(Img, colour_cals) + """ + calculate ratio of r, b to g + """ + r_g = np.mean(r_patchs/g_patchs) + b_g = np.mean(b_patchs/g_patchs) + Cam.log += '\n r : {:.4f} b : {:.4f}'.format(r_g, b_g) + """ + The curve tends to be better behaved in so-called hatspace. + R, B, G represent the individual channels. The colour curve is plotted in + r, b space, where: + r = R/G + b = B/G + This will be referred to as dehatspace... (sorry) + Hatspace is defined as: + r_hat = R/(R+B+G) + b_hat = B/(R+B+G) + To convert from dehatspace to hastpace (hat operation): + r_hat = r/(1+r+b) + b_hat = b/(1+r+b) + To convert from hatspace to dehatspace (dehat operation): + r = r_hat/(1-r_hat-b_hat) + b = b_hat/(1-r_hat-b_hat) + Proof is left as an excercise to the reader... + Throughout the code, r and b are sometimes referred to as r_g and b_g + as a reminder that they are ratios + """ + r_g_hat = r_g/(1+r_g+b_g) + b_g_hat = b_g/(1+r_g+b_g) + Cam.log += '\n r_hat : {:.4f} b_hat : {:.4f}'.format(r_g_hat, b_g_hat) + rbs_hat.append((r_g_hat, b_g_hat, Img.col)) + rb_raw.append((r_g, b_g)) + Cam.log += '\n' + + Cam.log += '\nFinished processing images' + """ + sort all lits simultaneously by r_hat + """ + rbs_zip = list(zip(rbs_hat, rb_raw)) + rbs_zip.sort(key=lambda x: x[0][0]) + rbs_hat, rb_raw = list(zip(*rbs_zip)) + """ + unzip tuples ready for processing + """ + rbs_hat = list(zip(*rbs_hat)) + rb_raw = list(zip(*rb_raw)) + """ + fit quadratic fit to r_g hat and b_g_hat + """ + a, b, c = np.polyfit(rbs_hat[0], rbs_hat[1], 2) + Cam.log += '\nFit quadratic curve in hatspace' + """ + the algorithm now approximates the shortest distance from each point to the + curve in dehatspace. Since the fit is done in hatspace, it is easier to + find the actual shortest distance in hatspace and use the projection back + into dehatspace as an overestimate. + The distance will be used for two things: + 1) In the case that colour temperature does not strictly decrease with + increasing r/g, the closest point to the line will be chosen out of an + increasing pair of colours. + + 2) To calculate transverse negative an dpositive, the maximum positive + and negative distance from the line are chosen. This benefits from the + overestimate as the transverse pos/neg are upper bound values. + """ + """ + define fit function + """ + def f(x): + return a*x**2 + b*x + c + """ + iterate over points (R, B are x and y coordinates of points) and calculate + distance to line in dehatspace + """ + dists = [] + for i, (R, B) in enumerate(zip(rbs_hat[0], rbs_hat[1])): + """ + define function to minimise as square distance between datapoint and + point on curve. Squaring is monotonic so minimising radius squared is + equivalent to minimising radius + """ + def f_min(x): + y = f(x) + return((x-R)**2+(y-B)**2) + """ + perform optimisation with scipy.optmisie.fmin + """ + x_hat = fmin(f_min, R, disp=0)[0] + y_hat = f(x_hat) + """ + dehat + """ + x = x_hat/(1-x_hat-y_hat) + y = y_hat/(1-x_hat-y_hat) + rr = R/(1-R-B) + bb = B/(1-R-B) + """ + calculate euclidean distance in dehatspace + """ + dist = ((x-rr)**2+(y-bb)**2)**0.5 + """ + return negative if point is below the fit curve + """ + if (x+y) > (rr+bb): + dist *= -1 + dists.append(dist) + Cam.log += '\nFound closest point on fit line to each point in dehatspace' + """ + calculate wiggle factors in awb. 10% added since this is an upper bound + """ + transverse_neg = - np.min(dists) * 1.1 + transverse_pos = np.max(dists) * 1.1 + Cam.log += '\nTransverse pos : {:.5f}'.format(transverse_pos) + Cam.log += '\nTransverse neg : {:.5f}'.format(transverse_neg) + """ + set minimum transverse wiggles to 0.1 . + Wiggle factors dictate how far off of the curve the algorithm searches. 0.1 + is a suitable minimum that gives better results for lighting conditions not + within calibration dataset. Anything less will generalise poorly. + """ + if transverse_pos < 0.01: + transverse_pos = 0.01 + Cam.log += '\nForced transverse pos to 0.01' + if transverse_neg < 0.01: + transverse_neg = 0.01 + Cam.log += '\nForced transverse neg to 0.01' + + """ + generate new b_hat values at each r_hat according to fit + """ + r_hat_fit = np.array(rbs_hat[0]) + b_hat_fit = a*r_hat_fit**2 + b*r_hat_fit + c + """ + transform from hatspace to dehatspace + """ + r_fit = r_hat_fit/(1-r_hat_fit-b_hat_fit) + b_fit = b_hat_fit/(1-r_hat_fit-b_hat_fit) + c_fit = np.round(rbs_hat[2], 0) + """ + round to 4dp + """ + r_fit = np.where((1000*r_fit) % 1 <= 0.05, r_fit+0.0001, r_fit) + r_fit = np.where((1000*r_fit) % 1 >= 0.95, r_fit-0.0001, r_fit) + b_fit = np.where((1000*b_fit) % 1 <= 0.05, b_fit+0.0001, b_fit) + b_fit = np.where((1000*b_fit) % 1 >= 0.95, b_fit-0.0001, b_fit) + r_fit = np.round(r_fit, 4) + b_fit = np.round(b_fit, 4) + """ + The following code ensures that colour temperature decreases with + increasing r/g + """ + """ + iterate backwards over list for easier indexing + """ + i = len(c_fit) - 1 + while i > 0: + if c_fit[i] > c_fit[i-1]: + Cam.log += '\nColour temperature increase found\n' + Cam.log += '{} K at r = {} to '.format(c_fit[i-1], r_fit[i-1]) + Cam.log += '{} K at r = {}'.format(c_fit[i], r_fit[i]) + """ + if colour temperature increases then discard point furthest from + the transformed fit (dehatspace) + """ + error_1 = abs(dists[i-1]) + error_2 = abs(dists[i]) + Cam.log += '\nDistances from fit:\n' + Cam.log += '{} K : {:.5f} , '.format(c_fit[i], error_1) + Cam.log += '{} K : {:.5f}'.format(c_fit[i-1], error_2) + """ + find bad index + note that in python false = 0 and true = 1 + """ + bad = i - (error_1 < error_2) + Cam.log += '\nPoint at {} K deleted as '.format(c_fit[bad]) + Cam.log += 'it is furthest from fit' + """ + delete bad point + """ + r_fit = np.delete(r_fit, bad) + b_fit = np.delete(b_fit, bad) + c_fit = np.delete(c_fit, bad).astype(np.uint16) + """ + note that if a point has been discarded then the length has decreased + by one, meaning that decreasing the index by one will reassess the kept + point against the next point. It is therefore possible, in theory, for + two adjacent points to be discarded, although probably rare + """ + i -= 1 + + """ + return formatted ct curve, ordered by increasing colour temperature + """ + ct_curve = list(np.array(list(zip(b_fit, r_fit, c_fit))).flatten())[::-1] + Cam.log += '\nFinal CT curve:' + for i in range(len(ct_curve)//3): + j = 3*i + Cam.log += '\n ct: {} '.format(ct_curve[j]) + Cam.log += ' r: {} '.format(ct_curve[j+1]) + Cam.log += ' b: {} '.format(ct_curve[j+2]) + + """ + plotting code for debug + """ + if plot: + x = np.linspace(np.min(rbs_hat[0]), np.max(rbs_hat[0]), 100) + y = a*x**2 + b*x + c + plt.subplot(2, 1, 1) + plt.title('hatspace') + plt.plot(rbs_hat[0], rbs_hat[1], ls='--', color='blue') + plt.plot(x, y, color='green', ls='-') + plt.scatter(rbs_hat[0], rbs_hat[1], color='red') + for i, ct in enumerate(rbs_hat[2]): + plt.annotate(str(ct), (rbs_hat[0][i], rbs_hat[1][i])) + plt.xlabel('$\\hat{r}$') + plt.ylabel('$\\hat{b}$') + """ + optional set axes equal to shortest distance so line really does + looks perpendicular and everybody is happy + """ + # ax = plt.gca() + # ax.set_aspect('equal') + plt.grid() + plt.subplot(2, 1, 2) + plt.title('dehatspace - indoors?') + plt.plot(r_fit, b_fit, color='blue') + plt.scatter(rb_raw[0], rb_raw[1], color='green') + plt.scatter(r_fit, b_fit, color='red') + for i, ct in enumerate(c_fit): + plt.annotate(str(ct), (r_fit[i], b_fit[i])) + plt.xlabel('$r$') + plt.ylabel('$b$') + """ + optional set axes equal to shortest distance so line really does + looks perpendicular and everybody is happy + """ + # ax = plt.gca() + # ax.set_aspect('equal') + plt.subplots_adjust(hspace=0.5) + plt.grid() + plt.show() + """ + end of plotting code + """ + return(ct_curve, np.round(transverse_pos, 5), np.round(transverse_neg, 5)) + + +""" +obtain greyscale patches and perform alsc colour correction +""" +def get_alsc_patches(Img, colour_cals, grey=True): + """ + get patch centre coordinates, image colour and the actual + patches for each channel, remembering to subtract blacklevel + If grey then only greyscale patches considered + """ + if grey: + cen_coords = Img.cen_coords[3::4] + col = Img.col + patches = [np.array(Img.patches[i]) for i in Img.order] + r_patchs = patches[0][3::4] - Img.blacklevel_16 + b_patchs = patches[3][3::4] - Img.blacklevel_16 + """ + note two green channels are averages + """ + g_patchs = (patches[1][3::4]+patches[2][3::4])/2 - Img.blacklevel_16 + else: + cen_coords = Img.cen_coords + col = Img.color + patches = [np.array(Img.patches[i]) for i in Img.order] + r_patchs = patches[0] - Img.blacklevel_16 + b_patchs = patches[3] - Img.blacklevel_16 + g_patchs = (patches[1]+patches[2])/2 - Img.blacklevel_16 + + if colour_cals is None: + return r_patchs, b_patchs, g_patchs + """ + find where image colour fits in alsc colour calibration tables + """ + cts = list(colour_cals.keys()) + pos = bisect_left(cts, col) + """ + if img colour is below minimum or above maximum alsc calibration colour, simply + pick extreme closest to img colour + """ + if pos % len(cts) == 0: + """ + this works because -0 = 0 = first and -1 = last index + """ + col_tabs = np.array(colour_cals[cts[-pos//len(cts)]]) + """ + else, perform linear interpolation between existing alsc colour + calibration tables + """ + else: + bef = cts[pos-1] + aft = cts[pos] + da = col-bef + db = aft-col + bef_tabs = np.array(colour_cals[bef]) + aft_tabs = np.array(colour_cals[aft]) + col_tabs = (bef_tabs*db + aft_tabs*da)/(da+db) + col_tabs = np.reshape(col_tabs, (2, 12, 16)) + """ + calculate dx, dy used to calculate alsc table + """ + w, h = Img.w/2, Img.h/2 + dx, dy = int(-(-(w-1)//16)), int(-(-(h-1)//12)) + """ + make list of pairs of gains for each patch by selecting the correct value + in alsc colour calibration table + """ + patch_gains = [] + for cen in cen_coords: + x, y = cen[0]//dx, cen[1]//dy + # We could probably do with some better spatial interpolation here? + col_gains = (col_tabs[0][y][x], col_tabs[1][y][x]) + patch_gains.append(col_gains) + + """ + multiply the r and b channels in each patch by the respective gain, finally + performing the alsc colour correction + """ + for i, gains in enumerate(patch_gains): + r_patchs[i] = r_patchs[i] * gains[0] + b_patchs[i] = b_patchs[i] * gains[1] + + """ + return greyscale patches, g channel and correct r, b channels + """ + return r_patchs, b_patchs, g_patchs diff --git a/utils/tuning/libtuning/ctt_ccm.py b/utils/tuning/libtuning/ctt_ccm.py new file mode 100644 index 00000000..2e87a667 --- /dev/null +++ b/utils/tuning/libtuning/ctt_ccm.py @@ -0,0 +1,408 @@ +# SPDX-License-Identifier: BSD-2-Clause +# +# Copyright (C) 2019, Raspberry Pi Ltd +# +# camera tuning tool for CCM (colour correction matrix) + +import logging + +import numpy as np +from scipy.optimize import minimize + +from . import ctt_colors as colors +from .image import Image +from .ctt_awb import get_alsc_patches +from .utils import visualise_macbeth_chart + +logger = logging.getLogger(__name__) + +""" +takes 8-bit macbeth chart values, degammas and returns 16 bit +""" + +''' +This program has many options from which to derive the color matrix from. +The first is average. This minimises the average delta E across all patches of +the macbeth chart. Testing across all cameras yeilded this as the most color +accurate and vivid. Other options are avalible however. +Maximum minimises the maximum Delta E of the patches. It iterates through till +a minimum maximum is found (so that there is +not one patch that deviates wildly.) +This yields generally good results but overall the colors are less accurate +Have a fiddle with maximum and see what you think. +The final option allows you to select the patches for which to average across. +This means that you can bias certain patches, for instance if you want the +reds to be more accurate. +''' + +matrix_selection_types = ["average", "maximum", "patches"] +typenum = 0 # select from array above, 0 = average, 1 = maximum, 2 = patches +test_patches = [1, 2, 5, 8, 9, 12, 14] + +''' +Enter patches to test for. Can also be entered twice if you +would like twice as much bias on one patch. +''' + + +def degamma(x): + x = x / ((2 ** 8) - 1) # takes 255 and scales it down to one + x = np.where(x < 0.04045, x / 12.92, ((x + 0.055) / 1.055) ** 2.4) + x = x * ((2 ** 16) - 1) # takes one and scales up to 65535, 16 bit color + return x + + +def gamma(x): + # Take 3 long array of color values and gamma them + return [((colour / 255) ** (1 / 2.4) * 1.055 - 0.055) * 255 for colour in x] + + +""" +FInds colour correction matrices for list of images +""" + + +def ccm(imgs, cal_cr_list, cal_cb_list): + global matrix_selection_types, typenum + """ + standard macbeth chart colour values + """ + m_rgb = np.array([ # these are in RGB + [116, 81, 67], # dark skin + [199, 147, 129], # light skin + [91, 122, 156], # blue sky + [90, 108, 64], # foliage + [130, 128, 176], # blue flower + [92, 190, 172], # bluish green + [224, 124, 47], # orange + [68, 91, 170], # purplish blue + [198, 82, 97], # moderate red + [94, 58, 106], # purple + [159, 189, 63], # yellow green + [230, 162, 39], # orange yellow + [35, 63, 147], # blue + [67, 149, 74], # green + [180, 49, 57], # red + [238, 198, 20], # yellow + [193, 84, 151], # magenta + [0, 136, 170], # cyan (goes out of gamut) + [245, 245, 243], # white 9.5 + [200, 202, 202], # neutral 8 + [161, 163, 163], # neutral 6.5 + [121, 121, 122], # neutral 5 + [82, 84, 86], # neutral 3.5 + [49, 49, 51] # black 2 + ]) + """ + convert reference colours from srgb to rgb + """ + m_srgb = degamma(m_rgb) # now in 16 bit color. + + # Produce array of LAB values for ideal color chart + m_lab = [colors.RGB_to_LAB(color / 256) for color in m_srgb] + + """ + reorder reference values to match how patches are ordered + """ + m_srgb = np.array([m_srgb[i::6] for i in range(6)]).reshape((24, 3)) + m_lab = np.array([m_lab[i::6] for i in range(6)]).reshape((24, 3)) + m_rgb = np.array([m_rgb[i::6] for i in range(6)]).reshape((24, 3)) + """ + reformat alsc correction tables or set colour_cals to None if alsc is + deactivated + """ + if cal_cr_list is None: + colour_cals = None + else: + colour_cals = {} + for cr, cb in zip(cal_cr_list, cal_cb_list): + cr_tab = cr['table'] + cb_tab = cb['table'] + """ + normalise tables so min value is 1 + """ + cr_tab = cr_tab / np.min(cr_tab) + cb_tab = cb_tab / np.min(cb_tab) + colour_cals[cr['ct']] = [cr_tab, cb_tab] + + """ + for each image, perform awb and alsc corrections. + Then calculate the colour correction matrix for that image, recording the + ccm and the colour tempertaure. + """ + ccm_tab = {} + for Img in imgs: + logger.info('Processing image: ' + Img.name) + """ + get macbeth patches with alsc applied if alsc enabled. + Note: if alsc is disabled then colour_cals will be set to None and no + the function will simply return the macbeth patches + """ + r, b, g = get_alsc_patches(Img, colour_cals, grey=False) + # 256 values for each patch of sRGB values + + """ + do awb + Note: awb is done by measuring the macbeth chart in the image, rather + than from the awb calibration. This is done so the awb will be perfect + and the ccm matrices will be more accurate. + """ + r_greys, b_greys, g_greys = r[3::4], b[3::4], g[3::4] + r_g = np.mean(r_greys / g_greys) + b_g = np.mean(b_greys / g_greys) + r = r / r_g + b = b / b_g + """ + normalise brightness wrt reference macbeth colours and then average + each channel for each patch + """ + gain = np.mean(m_srgb) / np.mean((r, g, b)) + logger.info(f'Gain with respect to standard colours: {gain:.3f}') + r = np.mean(gain * r, axis=1) + b = np.mean(gain * b, axis=1) + g = np.mean(gain * g, axis=1) + """ + calculate ccm matrix + """ + # ==== All of below should in sRGB ===## + sumde = 0 + ccm = do_ccm(r, g, b, m_srgb) + # This is the initial guess that our optimisation code works with. + original_ccm = ccm + r1 = ccm[0] + r2 = ccm[1] + g1 = ccm[3] + g2 = ccm[4] + b1 = ccm[6] + b2 = ccm[7] + ''' + COLOR MATRIX LOOKS AS BELOW + R1 R2 R3 Rval Outr + G1 G2 G3 * Gval = G + B1 B2 B3 Bval B + Will be optimising 6 elements and working out the third element using 1-r1-r2 = r3 + ''' + + x0 = [r1, r2, g1, g2, b1, b2] + ''' + We use our old CCM as the initial guess for the program to find the + optimised matrix + ''' + result = minimize(guess, x0, args=(r, g, b, m_lab), tol=0.01) + ''' + This produces a color matrix which has the lowest delta E possible, + based off the input data. Note it is impossible for this to reach + zero since the input data is imperfect + ''' + + [r1, r2, g1, g2, b1, b2] = result.x + # The new, optimised color correction matrix values + # This is the optimised Color Matrix (preserving greys by summing rows up to 1) + optimised_ccm = [r1, r2, (1 - r1 - r2), g1, g2, (1 - g1 - g2), b1, b2, (1 - b1 - b2)] + + logger.info(f'Optimized Matrix: {np.round(optimised_ccm, 4)}') + logger.info(f'Old Matrix: {np.round(ccm, 4)}') + + formatted_ccm = np.array(original_ccm).reshape((3, 3)) + + ''' + below is a whole load of code that then applies the latest color + matrix, and returns LAB values for color. This can then be used + to calculate the final delta E + ''' + optimised_ccm_rgb = [] # Original Color Corrected Matrix RGB / LAB + optimised_ccm_lab = [] + + formatted_optimised_ccm = np.array(optimised_ccm).reshape((3, 3)) + after_gamma_rgb = [] + after_gamma_lab = [] + + for RGB in zip(r, g, b): + ccm_applied_rgb = np.dot(formatted_ccm, (np.array(RGB) / 256)) + optimised_ccm_rgb.append(gamma(ccm_applied_rgb)) + optimised_ccm_lab.append(colors.RGB_to_LAB(ccm_applied_rgb)) + + optimised_ccm_applied_rgb = np.dot(formatted_optimised_ccm, np.array(RGB) / 256) + after_gamma_rgb.append(gamma(optimised_ccm_applied_rgb)) + after_gamma_lab.append(colors.RGB_to_LAB(optimised_ccm_applied_rgb)) + ''' + Gamma After RGB / LAB - not used in calculations, only used for visualisation + We now want to spit out some data that shows + how the optimisation has improved the color matrices + ''' + logger.info("Here are the Improvements") + + # CALCULATE WORST CASE delta e + old_worst_delta_e = 0 + before_average = transform_and_evaluate(formatted_ccm, r, g, b, m_lab) + new_worst_delta_e = 0 + after_average = transform_and_evaluate(formatted_optimised_ccm, r, g, b, m_lab) + for i in range(24): + old_delta_e = deltae(optimised_ccm_lab[i], m_lab[i]) # Current Old Delta E + new_delta_e = deltae(after_gamma_lab[i], m_lab[i]) # Current New Delta E + if old_delta_e > old_worst_delta_e: + old_worst_delta_e = old_delta_e + if new_delta_e > new_worst_delta_e: + new_worst_delta_e = new_delta_e + + logger.info(f'delta E optimized: average: {after_average:.2f} max:{new_worst_delta_e:.2f}') + logger.info(f'delta E old: average: {before_average:.2f} max:{old_worst_delta_e:.2f}') + + visualise_macbeth_chart(m_rgb, optimised_ccm_rgb, after_gamma_rgb, str(Img.color) + str(matrix_selection_types[typenum])) + ''' + The program will also save some visualisations of improvements. + Very pretty to look at. Top rectangle is ideal, Left square is + before optimisation, right square is after. + ''' + + """ + if a ccm has already been calculated for that temperature then don't + overwrite but save both. They will then be averaged later on + """ # Now going to use optimised color matrix, optimised_ccm + if Img.color in ccm_tab.keys(): + ccm_tab[Img.color].append(optimised_ccm) + else: + ccm_tab[Img.color] = [optimised_ccm] + + logger.info('Finished processing images') + """ + average any ccms that share a colour temperature + """ + for k, v in ccm_tab.items(): + tab = np.mean(v, axis=0) + tab = np.where((10000 * tab) % 1 <= 0.05, tab + 0.00001, tab) + tab = np.where((10000 * tab) % 1 >= 0.95, tab - 0.00001, tab) + ccm_tab[k] = list(np.round(tab, 5)) + logger.info(f'Matrix calculated for colour temperature of {k} K') + + """ + return all ccms with respective colour temperature in the correct format, + sorted by their colour temperature + """ + sorted_ccms = sorted(ccm_tab.items(), key=lambda kv: kv[0]) + ccms = [] + for i in sorted_ccms: + ccms.append({ + 'ct': i[0], + 'ccm': i[1] + }) + return ccms + + +def guess(x0, r, g, b, m_lab): # provides a method of numerical feedback for the optimisation code + [r1, r2, g1, g2, b1, b2] = x0 + ccm = np.array([r1, r2, (1 - r1 - r2), + g1, g2, (1 - g1 - g2), + b1, b2, (1 - b1 - b2)]).reshape((3, 3)) # format the matrix correctly + return transform_and_evaluate(ccm, r, g, b, m_lab) + + +def transform_and_evaluate(ccm, r, g, b, m_lab): # Transforms colors to LAB and applies the correction matrix + # create list of matrix changed colors + realrgb = [] + for RGB in zip(r, g, b): + rgb_post_ccm = np.dot(ccm, np.array(RGB) / 256) # This is RGB values after the color correction matrix has been applied + realrgb.append(colors.RGB_to_LAB(rgb_post_ccm)) + # now compare that with m_lab and return numeric result, averaged for each patch + return (sumde(realrgb, m_lab) / 24) # returns an average result of delta E + + +def sumde(listA, listB): + global typenum, test_patches + sumde = 0 + maxde = 0 + patchde = [] # Create array of the delta E values for each patch. useful for optimisation of certain patches + for listA_item, listB_item in zip(listA, listB): + if maxde < (deltae(listA_item, listB_item)): + maxde = deltae(listA_item, listB_item) + patchde.append(deltae(listA_item, listB_item)) + sumde += deltae(listA_item, listB_item) + ''' + The different options specified at the start allow for + the maximum to be returned, average or specific patches + ''' + if typenum == 0: + return sumde + if typenum == 1: + return maxde + if typenum == 2: + output = sum([patchde[test_patch] for test_patch in test_patches]) + # Selects only certain patches and returns the output for them + return output + + +""" +calculates the ccm for an individual image. +ccms are calculated in rgb space, and are fit by hand. Although it is a 3x3 +matrix, each row must add up to 1 in order to conserve greyness, simplifying +calculation. +The initial CCM is calculated in RGB, and then optimised in LAB color space +This simplifies the initial calculation but then gets us the accuracy of +using LAB color space. +""" + + +def do_ccm(r, g, b, m_srgb): + rb = r-b + gb = g-b + rb_2s = (rb * rb) + rb_gbs = (rb * gb) + gb_2s = (gb * gb) + + r_rbs = rb * (m_srgb[..., 0] - b) + r_gbs = gb * (m_srgb[..., 0] - b) + g_rbs = rb * (m_srgb[..., 1] - b) + g_gbs = gb * (m_srgb[..., 1] - b) + b_rbs = rb * (m_srgb[..., 2] - b) + b_gbs = gb * (m_srgb[..., 2] - b) + + """ + Obtain least squares fit + """ + rb_2 = np.sum(rb_2s) + gb_2 = np.sum(gb_2s) + rb_gb = np.sum(rb_gbs) + r_rb = np.sum(r_rbs) + r_gb = np.sum(r_gbs) + g_rb = np.sum(g_rbs) + g_gb = np.sum(g_gbs) + b_rb = np.sum(b_rbs) + b_gb = np.sum(b_gbs) + + det = rb_2 * gb_2 - rb_gb * rb_gb + + """ + Raise error if matrix is singular... + This shouldn't really happen with real data but if it does just take new + pictures and try again, not much else to be done unfortunately... + """ + if det < 0.001: + raise ArithmeticError + + r_a = (gb_2 * r_rb - rb_gb * r_gb) / det + r_b = (rb_2 * r_gb - rb_gb * r_rb) / det + """ + Last row can be calculated by knowing the sum must be 1 + """ + r_c = 1 - r_a - r_b + + g_a = (gb_2 * g_rb - rb_gb * g_gb) / det + g_b = (rb_2 * g_gb - rb_gb * g_rb) / det + g_c = 1 - g_a - g_b + + b_a = (gb_2 * b_rb - rb_gb * b_gb) / det + b_b = (rb_2 * b_gb - rb_gb * b_rb) / det + b_c = 1 - b_a - b_b + + """ + format ccm + """ + ccm = [r_a, r_b, r_c, g_a, g_b, g_c, b_a, b_b, b_c] + + return ccm + + +def deltae(colorA, colorB): + return ((colorA[0] - colorB[0]) ** 2 + (colorA[1] - colorB[1]) ** 2 + (colorA[2] - colorB[2]) ** 2) ** 0.5 + # return ((colorA[1]-colorB[1]) * * 2 + (colorA[2]-colorB[2]) * * 2) * * 0.5 + # UNCOMMENT IF YOU WANT TO NEGLECT LUMINANCE FROM CALCULATION OF DELTA E diff --git a/utils/tuning/libtuning/ctt_colors.py b/utils/tuning/libtuning/ctt_colors.py new file mode 100644 index 00000000..cb4d236b --- /dev/null +++ b/utils/tuning/libtuning/ctt_colors.py @@ -0,0 +1,30 @@ +# Program to convert from RGB to LAB color space +def RGB_to_LAB(RGB): # where RGB is a 1x3 array. e.g RGB = [100, 255, 230] + num = 0 + XYZ = [0, 0, 0] + # converted all the three R, G, B to X, Y, Z + X = RGB[0] * 0.4124 + RGB[1] * 0.3576 + RGB[2] * 0.1805 + Y = RGB[0] * 0.2126 + RGB[1] * 0.7152 + RGB[2] * 0.0722 + Z = RGB[0] * 0.0193 + RGB[1] * 0.1192 + RGB[2] * 0.9505 + + XYZ[0] = X / 255 * 100 + XYZ[1] = Y / 255 * 100 # XYZ Must be in range 0 -> 100, so scale down from 255 + XYZ[2] = Z / 255 * 100 + XYZ[0] = XYZ[0] / 95.047 # ref_X = 95.047 Observer= 2°, Illuminant= D65 + XYZ[1] = XYZ[1] / 100.0 # ref_Y = 100.000 + XYZ[2] = XYZ[2] / 108.883 # ref_Z = 108.883 + num = 0 + for value in XYZ: + if value > 0.008856: + value = value ** (0.3333333333333333) + else: + value = (7.787 * value) + (16 / 116) + XYZ[num] = value + num = num + 1 + + # L, A, B, values calculated below + L = (116 * XYZ[1]) - 16 + a = 500 * (XYZ[0] - XYZ[1]) + b = 200 * (XYZ[1] - XYZ[2]) + + return [L, a, b] diff --git a/utils/tuning/libtuning/ctt_ransac.py b/utils/tuning/libtuning/ctt_ransac.py new file mode 100644 index 00000000..01bba302 --- /dev/null +++ b/utils/tuning/libtuning/ctt_ransac.py @@ -0,0 +1,71 @@ +# SPDX-License-Identifier: BSD-2-Clause +# +# Copyright (C) 2019, Raspberry Pi Ltd +# +# camera tuning tool RANSAC selector for Macbeth chart locator + +import numpy as np + +scale = 2 + + +""" +constructs normalised macbeth chart corners for ransac algorithm +""" +def get_square_verts(c_err=0.05, scale=scale): + """ + define macbeth chart corners + """ + b_bord_x, b_bord_y = scale*8.5, scale*13 + s_bord = 6*scale + side = 41*scale + x_max = side*6 + 5*s_bord + 2*b_bord_x + y_max = side*4 + 3*s_bord + 2*b_bord_y + c1 = (0, 0) + c2 = (0, y_max) + c3 = (x_max, y_max) + c4 = (x_max, 0) + mac_norm = np.array((c1, c2, c3, c4), np.float32) + mac_norm = np.array([mac_norm]) + + square_verts = [] + square_0 = np.array(((0, 0), (0, side), + (side, side), (side, 0)), np.float32) + offset_0 = np.array((b_bord_x, b_bord_y), np.float32) + c_off = side * c_err + offset_cont = np.array(((c_off, c_off), (c_off, -c_off), + (-c_off, -c_off), (-c_off, c_off)), np.float32) + square_0 += offset_0 + square_0 += offset_cont + """ + define macbeth square corners + """ + for i in range(6): + shift_i = np.array(((i*side, 0), (i*side, 0), + (i*side, 0), (i*side, 0)), np.float32) + shift_bord = np.array(((i*s_bord, 0), (i*s_bord, 0), + (i*s_bord, 0), (i*s_bord, 0)), np.float32) + square_i = square_0 + shift_i + shift_bord + for j in range(4): + shift_j = np.array(((0, j*side), (0, j*side), + (0, j*side), (0, j*side)), np.float32) + shift_bord = np.array(((0, j*s_bord), + (0, j*s_bord), (0, j*s_bord), + (0, j*s_bord)), np.float32) + square_j = square_i + shift_j + shift_bord + square_verts.append(square_j) + # print('square_verts') + # print(square_verts) + return np.array(square_verts, np.float32), mac_norm + + +def get_square_centres(c_err=0.05, scale=scale): + """ + define macbeth square centres + """ + verts, mac_norm = get_square_verts(c_err, scale=scale) + + centres = np.mean(verts, axis=1) + # print('centres') + # print(centres) + return np.array(centres, np.float32) diff --git a/utils/tuning/libtuning/generators/yaml_output.py b/utils/tuning/libtuning/generators/yaml_output.py index 8f22d386..c490081d 100644 --- a/utils/tuning/libtuning/generators/yaml_output.py +++ b/utils/tuning/libtuning/generators/yaml_output.py @@ -9,8 +9,9 @@ from .generator import Generator from numbers import Number from pathlib import Path -import libtuning.utils as utils +import logging +logger = logging.getLogger(__name__) class YamlOutput(Generator): def __init__(self): @@ -106,13 +107,16 @@ class YamlOutput(Generator): ] for module in output_order: + if module not in output_dict: + continue + out_lines.append(f' - {module.out_name}:') if len(output_dict[module]) == 0: continue if not isinstance(output_dict[module], dict): - utils.eprint(f'Error: Output of {module.type} is not a dictionary') + logger.error(f'Error: Output of {module.type} is not a dictionary') continue lines = self._stringify_dict(output_dict[module]) diff --git a/utils/tuning/libtuning/image.py b/utils/tuning/libtuning/image.py index e2181b11..c8911a0f 100644 --- a/utils/tuning/libtuning/image.py +++ b/utils/tuning/libtuning/image.py @@ -13,6 +13,9 @@ import re import libtuning as lt import libtuning.utils as utils +import logging + +logger = logging.getLogger(__name__) class Image: @@ -21,17 +24,18 @@ class Image: self.lsc_only = False self.color = -1 self.lux = -1 + self.macbeth = None try: self._load_metadata_exif() except Exception as e: - utils.eprint(f'Failed to load metadata from {self.path}: {e}') + logger.error(f'Failed to load metadata from {self.path}: {e}') raise e try: self._read_image_dng() except Exception as e: - utils.eprint(f'Failed to load image data from {self.path}: {e}') + logger.error(f'Failed to load image data from {self.path}: {e}') raise e @property @@ -79,7 +83,7 @@ class Image: # is R, then G, then G, then B. bayer_case = { '0 1 1 2': (lt.Color.R, lt.Color.GR, lt.Color.GB, lt.Color.B), - '1 2 0 1': (lt.Color.GB, lt.Color.R, lt.Color.B, lt.Color.GR), + '1 2 0 1': (lt.Color.GB, lt.Color.B, lt.Color.R, lt.Color.GR), '2 1 1 0': (lt.Color.B, lt.Color.GB, lt.Color.GR, lt.Color.R), '1 0 2 1': (lt.Color.GR, lt.Color.R, lt.Color.B, lt.Color.GB) } diff --git a/utils/tuning/libtuning/libtuning.py b/utils/tuning/libtuning/libtuning.py index 5e22288d..bac57323 100644 --- a/utils/tuning/libtuning/libtuning.py +++ b/utils/tuning/libtuning/libtuning.py @@ -5,13 +5,14 @@ # An infrastructure for camera tuning tools import argparse +import logging import libtuning as lt import libtuning.utils as utils -from libtuning.utils import eprint from enum import Enum, IntEnum +logger = logging.getLogger(__name__) class Color(IntEnum): R = 0 @@ -94,7 +95,10 @@ class Tuner(object): self.output = {} def add(self, module): - self.modules.append(module) + if isinstance(module, list): + self.modules.extend(module) + else: + self.modules.append(module) def set_input_parser(self, parser): self.parser = parser @@ -112,10 +116,10 @@ class Tuner(object): for module_type in output_order: modules = [module for module in self.modules if module.type == module_type.type] if len(modules) > 1: - eprint(f'Multiple modules found for module type "{module_type.type}"') + logger.error(f'Multiple modules found for module type "{module_type.type}"') return False if len(modules) < 1: - eprint(f'No module found for module type "{module_type.type}"') + logger.error(f'No module found for module type "{module_type.type}"') return False self.output_order.append(modules[0]) @@ -124,19 +128,19 @@ class Tuner(object): # \todo Validate parser and generator at Tuner construction time? def _validate_settings(self): if self.parser is None: - eprint('Missing parser') + logger.error('Missing parser') return False if self.generator is None: - eprint('Missing generator') + logger.error('Missing generator') return False if len(self.modules) == 0: - eprint('No modules added') + logger.error('No modules added') return False if len(self.output_order) != len(self.modules): - eprint('Number of outputs does not match number of modules') + logger.error('Number of outputs does not match number of modules') return False return True @@ -183,7 +187,7 @@ class Tuner(object): for module in self.modules: if not module.validate_config(self.config): - eprint(f'Config is invalid for module {module.type}') + logger.error(f'Config is invalid for module {module.type}') return -1 has_lsc = any(isinstance(m, lt.modules.lsc.LSC) for m in self.modules) @@ -192,15 +196,15 @@ class Tuner(object): images = utils.load_images(args.input, self.config, not has_only_lsc, has_lsc) if images is None or len(images) == 0: - eprint(f'No images were found, or able to load') + logger.error(f'No images were found, or able to load') return -1 # Do the tuning for module in self.modules: out = module.process(self.config, images, self.output) if out is None: - eprint(f'Module {module.name} failed to process, aborting') - break + logger.warning(f'Module {module.hr_name} failed to process...') + continue self.output[module] = out self.generator.write(args.output, self.output, self.output_order) diff --git a/utils/tuning/libtuning/macbeth.py b/utils/tuning/libtuning/macbeth.py index e1182464..4a2006b0 100644 --- a/utils/tuning/libtuning/macbeth.py +++ b/utils/tuning/libtuning/macbeth.py @@ -1,6 +1,7 @@ # SPDX-License-Identifier: BSD-2-Clause # # Copyright (C) 2019, Raspberry Pi Ltd +# Copyright (C) 2024, Ideas on Board Oy # # Locate and extract Macbeth charts from images # (Copied from: ctt_macbeth_locator.py) @@ -11,8 +12,18 @@ import cv2 import os from pathlib import Path import numpy as np +import warnings +import logging +from sklearn import cluster as cluster -from libtuning.image import Image +from .ctt_ransac import get_square_verts, get_square_centres +from .image import Image + +logger = logging.getLogger(__name__) + + +class MacbethError(Exception): + pass # Reshape image to fixed width without distorting returns image and scale @@ -369,7 +380,9 @@ def get_macbeth_chart(img, ref_data): # Catch macbeth errors and continue with code except MacbethError as error: - eprint(error) + # \todo: This happens so many times in a normal run, that it shadows + # all the relevant output + # logger.warning(error) return (0, None, None, False) @@ -403,10 +416,15 @@ def find_macbeth(img, mac_config): # nothing more is tried as this is a high enough confidence to ensure # reliable macbeth square centre placement. + # Keep a list that will include this and any brightened up versions of + # the image for reuse. + all_images = [img] + for brightness in [2, 4]: if cor >= 0.75: break img_br = cv2.convertScaleAbs(img, alpha=brightness, beta=0) + all_images.append(img_br) cor_b, mac_b, coords_b, ret_b = get_macbeth_chart(img_br, ref_data) if cor_b > cor: cor, mac, coords, ret = cor_b, mac_b, coords_b, ret_b @@ -456,23 +474,24 @@ def find_macbeth(img, mac_config): w_inc = int(w * pair['inc']) h_inc = int(h * pair['inc']) - loop = ((1 - pair['sel']) / pair['inc']) + 1 + loop = int(((1 - pair['sel']) / pair['inc']) + 1) # For each subselection, look for a macbeth chart - for i in range(loop): - for j in range(loop): - w_s, h_s = i * w_inc, j * h_inc - img_sel = img[w_s:w_s + w_sel, h_s:h_s + h_sel] - cor_ij, mac_ij, coords_ij, ret_ij = get_macbeth_chart(img_sel, ref_data) - - # If the correlation is better than the best then record the - # scale and current subselection at which macbeth chart was - # found. Also record the coordinates, macbeth chart and message. - if cor_ij > cor: - cor = cor_ij - mac, coords, ret = mac_ij, coords_ij, ret_ij - ii, jj = i, j - w_best, h_best = w_inc, h_inc - d_best = index + 1 + for img_br in all_images: + for i in range(loop): + for j in range(loop): + w_s, h_s = i * w_inc, j * h_inc + img_sel = img_br[w_s:w_s + w_sel, h_s:h_s + h_sel] + cor_ij, mac_ij, coords_ij, ret_ij = get_macbeth_chart(img_sel, ref_data) + + # If the correlation is better than the best then record the + # scale and current subselection at which macbeth chart was + # found. Also record the coordinates, macbeth chart and message. + if cor_ij > cor: + cor = cor_ij + mac, coords, ret = mac_ij, coords_ij, ret_ij + ii, jj = i, j + w_best, h_best = w_inc, h_inc + d_best = index + 1 # Transform coordinates from subselection to original image if ii != -1: @@ -486,7 +505,7 @@ def find_macbeth(img, mac_config): coords_fit = coords if cor < 0.75: - eprint(f'Warning: Low confidence {cor:.3f} for macbeth chart in {img.path.name}') + logger.warning(f'Low confidence {cor:.3f} for macbeth chart') if show: draw_macbeth_results(img, coords_fit) @@ -499,18 +518,20 @@ def locate_macbeth(image: Image, config: dict): av_chan = (np.mean(np.array(image.channels), axis=0) / (2**16)) av_val = np.mean(av_chan) if av_val < image.blacklevel_16 / (2**16) + 1 / 64: - eprint(f'Image {image.path.name} too dark') + logger.warning(f'Image {image.path.name} too dark') return None macbeth = find_macbeth(av_chan, config['general']['macbeth']) if macbeth is None: - eprint(f'No macbeth chart found in {image.path.name}') + logger.warning(f'No macbeth chart found in {image.path.name}') return None mac_cen_coords = macbeth[1] if not image.get_patches(mac_cen_coords): - eprint(f'Macbeth patches have saturated in {image.path.name}') + logger.warning(f'Macbeth patches have saturated in {image.path.name}') return None + image.macbeth = macbeth + return macbeth diff --git a/utils/tuning/libtuning/macbeth_ref.pgm b/utils/tuning/libtuning/macbeth_ref.pgm index 37897140..089ea91f 100644 --- a/utils/tuning/libtuning/macbeth_ref.pgm +++ b/utils/tuning/libtuning/macbeth_ref.pgm @@ -1,5 +1,5 @@ -# SPDX-License-Identifier: BSD-2-Clause P5 +# SPDX-License-Identifier: BSD-2-Clause # Reference macbeth chart 120 80 255 diff --git a/utils/tuning/libtuning/modules/agc/rkisp1.py b/utils/tuning/libtuning/modules/agc/rkisp1.py index 19a5555b..7147028a 100644 --- a/utils/tuning/libtuning/modules/agc/rkisp1.py +++ b/utils/tuning/libtuning/modules/agc/rkisp1.py @@ -64,7 +64,7 @@ class AGCRkISP1(AGC): return {'ConstraintNormal': normal, 'ConstraintHighlight': highlight} def _generate_y_target(self) -> list: - return 0.16 + return 0.5 def process(self, config: dict, images: list, outputs: dict) -> dict: output = {} diff --git a/utils/tuning/libtuning/modules/ccm/__init__.py b/utils/tuning/libtuning/modules/ccm/__init__.py new file mode 100644 index 00000000..322602af --- /dev/null +++ b/utils/tuning/libtuning/modules/ccm/__init__.py @@ -0,0 +1,6 @@ +# SPDX-License-Identifier: GPL-2.0-or-later +# +# Copyright (C) 2024, Paul Elder <paul.elder@ideasonboard.com> + +from libtuning.modules.ccm.ccm import CCM +from libtuning.modules.ccm.rkisp1 import CCMRkISP1 diff --git a/utils/tuning/libtuning/modules/ccm/ccm.py b/utils/tuning/libtuning/modules/ccm/ccm.py new file mode 100644 index 00000000..18702f8d --- /dev/null +++ b/utils/tuning/libtuning/modules/ccm/ccm.py @@ -0,0 +1,41 @@ +# SPDX-License-Identifier: GPL-2.0-or-later +# +# Copyright (C) 2024, Paul Elder <paul.elder@ideasonboard.com> +# Copyright (C) 2024, Ideas on Board +# +# Base Ccm tuning module + +from ..module import Module + +from libtuning.ctt_ccm import ccm +import logging + +logger = logging.getLogger(__name__) + + +class CCM(Module): + type = 'ccm' + hr_name = 'CCM (Base)' + out_name = 'GenericCCM' + + def __init__(self, debug: list): + super().__init__() + + self.debug = debug + + def do_calibration(self, images): + logger.info('Starting CCM calibration') + + imgs = [img for img in images if img.macbeth is not None] + + # todo: Take LSC calibration results into account. + cal_cr_list = None + cal_cb_list = None + + try: + ccms = ccm(imgs, cal_cr_list, cal_cb_list) + except ArithmeticError: + logger.error('CCM calibration failed') + return None + + return ccms diff --git a/utils/tuning/libtuning/modules/ccm/rkisp1.py b/utils/tuning/libtuning/modules/ccm/rkisp1.py new file mode 100644 index 00000000..be0252d9 --- /dev/null +++ b/utils/tuning/libtuning/modules/ccm/rkisp1.py @@ -0,0 +1,28 @@ +# SPDX-License-Identifier: GPL-2.0-or-later +# +# Copyright (C) 2024, Paul Elder <paul.elder@ideasonboard.com> +# Copyright (C) 2024, Ideas on Board +# +# Ccm module for tuning rkisp1 + +from .ccm import CCM + + +class CCMRkISP1(CCM): + hr_name = 'Crosstalk Correction (RkISP1)' + out_name = 'Ccm' + + def __init__(self, **kwargs): + super().__init__(**kwargs) + + # We don't need anything from the config file. + def validate_config(self, config: dict) -> bool: + return True + + def process(self, config: dict, images: list, outputs: dict) -> dict: + output = {} + + ccms = self.do_calibration(images) + output['ccms'] = ccms + + return output diff --git a/utils/tuning/libtuning/modules/lsc/lsc.py b/utils/tuning/libtuning/modules/lsc/lsc.py index 344a07a3..e0ca22eb 100644 --- a/utils/tuning/libtuning/modules/lsc/lsc.py +++ b/utils/tuning/libtuning/modules/lsc/lsc.py @@ -59,7 +59,10 @@ class LSC(Module): def _lsc_single_channel(self, channel: np.array, image: lt.Image, green_grid: np.array = None): grid = self._get_grid(channel, image.w, image.h) - grid -= image.blacklevel_16 + # Clamp the values to a small positive, so that the following 1/grid + # doesn't produce negative results. + grid = np.maximum(grid - image.blacklevel_16, 0.1) + if green_grid is None: table = np.reshape(1 / grid, self.sector_shape[::-1]) else: diff --git a/utils/tuning/libtuning/modules/lsc/raspberrypi.py b/utils/tuning/libtuning/modules/lsc/raspberrypi.py index f19c7163..99bc4fe6 100644 --- a/utils/tuning/libtuning/modules/lsc/raspberrypi.py +++ b/utils/tuning/libtuning/modules/lsc/raspberrypi.py @@ -12,7 +12,9 @@ import libtuning.utils as utils from numbers import Number import numpy as np +import logging +logger = logging.getLogger(__name__) class ALSCRaspberryPi(LSC): # Override the type name so that the parser can match the entry in the @@ -35,7 +37,7 @@ class ALSCRaspberryPi(LSC): def validate_config(self, config: dict) -> bool: if self not in config: - utils.eprint(f'{self.type} not in config') + logger.error(f'{self.type} not in config') return False valid = True @@ -46,14 +48,14 @@ class ALSCRaspberryPi(LSC): color_key = self.do_color.name if lum_key not in conf and self.luminance_strength.required: - utils.eprint(f'{lum_key} is not in config') + logger.error(f'{lum_key} is not in config') valid = False if lum_key in conf and (conf[lum_key] < 0 or conf[lum_key] > 1): - utils.eprint(f'Warning: {lum_key} is not in range [0, 1]; defaulting to 0.5') + logger.warning(f'{lum_key} is not in range [0, 1]; defaulting to 0.5') if color_key not in conf and self.do_color.required: - utils.eprint(f'{color_key} is not in config') + logger.error(f'{color_key} is not in config') valid = False return valid @@ -235,7 +237,7 @@ class ALSCRaspberryPi(LSC): if count == 1: output['sigma'] = 0.005 output['sigma_Cb'] = 0.005 - utils.eprint('Warning: Only one alsc calibration found; standard sigmas used for adaptive algorithm.') + logger.warning('Only one alsc calibration found; standard sigmas used for adaptive algorithm.') return output # Obtain worst-case scenario residual sigmas diff --git a/utils/tuning/libtuning/modules/lsc/rkisp1.py b/utils/tuning/libtuning/modules/lsc/rkisp1.py index 20406e43..c02b2306 100644 --- a/utils/tuning/libtuning/modules/lsc/rkisp1.py +++ b/utils/tuning/libtuning/modules/lsc/rkisp1.py @@ -33,13 +33,13 @@ class LSCRkISP1(LSC): # table, flattened array of (blue's) green calibration table def _do_single_lsc(self, image: lt.Image): - cgr, gr = self._lsc_single_channel(image.channels[lt.Color.GR], image) - cgb, gb = self._lsc_single_channel(image.channels[lt.Color.GB], image) - - # \todo Should these ratio against the average of both greens or just - # each green like we've done here? - cr, _ = self._lsc_single_channel(image.channels[lt.Color.R], image, gr) - cb, _ = self._lsc_single_channel(image.channels[lt.Color.B], image, gb) + # Perform LSC on each colour channel independently. A future enhancement + # worth investigating would be splitting the luminance and chrominance + # LSC as done by Raspberry Pi. + cgr, _ = self._lsc_single_channel(image.channels[lt.Color.GR], image) + cgb, _ = self._lsc_single_channel(image.channels[lt.Color.GB], image) + cr, _ = self._lsc_single_channel(image.channels[lt.Color.R], image) + cb, _ = self._lsc_single_channel(image.channels[lt.Color.B], image) return image.color, cr.flatten(), cb.flatten(), cgr.flatten(), cgb.flatten() @@ -80,7 +80,8 @@ class LSCRkISP1(LSC): tables = [] for lis in [list_cr, list_cgr, list_cgb, list_cb]: table = np.mean(lis[indices], axis=0) - table = output_map_func((1, 3.999), (1024, 4095), table) + table = output_map_func((1, 4), (1024, 4096), table) + table = np.clip(table, 1024, 4095) table = np.round(table).astype('int32').tolist() tables.append(table) @@ -106,6 +107,9 @@ class LSCRkISP1(LSC): output['sets'] = self._do_all_lsc(images) + if len(output['sets']) == 0: + return None + # \todo Validate images from greyscale camera and force grescale mode # \todo Debug functionality diff --git a/utils/tuning/libtuning/modules/static.py b/utils/tuning/libtuning/modules/static.py new file mode 100644 index 00000000..4d0f7e18 --- /dev/null +++ b/utils/tuning/libtuning/modules/static.py @@ -0,0 +1,24 @@ +# SPDX-License-Identifier: GPL-2.0-or-later +# +# Copyright (C) 2024, Ideas on Board +# +# Module implementation for static data + +from .module import Module + + +# This module can be used in cases where the tuning file should contain +# static data. +class StaticModule(Module): + def __init__(self, out_name: str, output: dict = {}): + super().__init__() + self.out_name = out_name + self.hr_name = f'Static {out_name}' + self.type = f'static_{out_name}' + self.output = output + + def validate_config(self, config: dict) -> bool: + return True + + def process(self, config: dict, images: list, outputs: dict) -> dict: + return self.output diff --git a/utils/tuning/libtuning/parsers/yaml_parser.py b/utils/tuning/libtuning/parsers/yaml_parser.py index 244db24d..1fa6b7a8 100644 --- a/utils/tuning/libtuning/parsers/yaml_parser.py +++ b/utils/tuning/libtuning/parsers/yaml_parser.py @@ -5,13 +5,16 @@ # Parser for YAML format config file from .parser import Parser +import yaml class YamlParser(Parser): def __init__(self): super().__init__() - # \todo Implement this (it's fine for now as we don't need a config for - # rkisp1 LSC, which is the only user of this so far) def parse(self, config_file: str, modules: list) -> (dict, list): - return {}, [] + # Dummy implementation that just reads the file + with open(config_file, 'r') as f: + config = yaml.safe_load(f) + + return config, [] diff --git a/utils/tuning/libtuning/utils.py b/utils/tuning/libtuning/utils.py index 1e8128ea..e35cf409 100644 --- a/utils/tuning/libtuning/utils.py +++ b/utils/tuning/libtuning/utils.py @@ -5,6 +5,7 @@ # # Utilities for libtuning +import cv2 import decimal import math import numpy as np @@ -12,16 +13,15 @@ import os from pathlib import Path import re import sys +import logging import libtuning as lt from libtuning.image import Image -from libtuning.macbeth import locate_macbeth - -# Utility functions +from .macbeth import locate_macbeth +logger = logging.getLogger(__name__) -def eprint(*args, **kwargs): - print(*args, file=sys.stderr, **kwargs) +# Utility functions def get_module_by_type_name(modules, name): @@ -43,16 +43,30 @@ def _list_image_files(directory): def _parse_image_filename(fn: Path): - result = re.search(r'^(alsc_)?(\d+)[kK]_(\d+)?[lLuU]?.\w{3,4}$', fn.name) - if result is None: - eprint(f'The file name of {fn.name} is incorrectly formatted') - return None, None, None + lsc_only = False + color_temperature = None + lux = None + + parts = fn.stem.split('_') + for part in parts: + if part == 'alsc': + lsc_only = True + continue + r = re.match(r'(\d+)[kK]', part) + if r: + color_temperature = int(r.group(1)) + continue + r = re.match(r'(\d+)[lLuU]', part) + if r: + lux = int(r.group(1)) + + if color_temperature is None: + logger.error(f'The file name of "{fn.name}" does not contain a color temperature') - color = int(result.group(2)) - lsc_only = result.group(1) is not None - lux = None if lsc_only else int(result.group(3)) + if lux is None and lsc_only is False: + logger.error(f'The file name of "{fn.name}" must either contain alsc or a lux level') - return color, lux, lsc_only + return color_temperature, lux, lsc_only # \todo Implement this from check_imgs() in ctt.py @@ -72,30 +86,34 @@ def _validate_images(images): def load_images(input_dir: str, config: dict, load_nonlsc: bool, load_lsc: bool) -> list: files = _list_image_files(input_dir) if len(files) == 0: - eprint(f'No images found in {input_dir}') + logger.error(f'No images found in {input_dir}') return None images = [] for f in files: color, lux, lsc_only = _parse_image_filename(f) + if color is None: + logger.warning(f'Ignoring "{f.name}" as it has no associated color temperature') continue + logger.info(f'Process image "{f.name}" (color={color}, lux={lux}, lsc_only={lsc_only})') + # Skip lsc image if we don't need it if lsc_only and not load_lsc: - eprint(f'Skipping {f.name} as this tuner has no LSC module') + logger.warning(f'Skipping {f.name} as this tuner has no LSC module') continue # Skip non-lsc image if we don't need it if not lsc_only and not load_nonlsc: - eprint(f'Skipping {f.name} as this tuner only has an LSC module') + logger.warning(f'Skipping {f.name} as this tuner only has an LSC module') continue # Load image try: image = Image(f) except Exception as e: - eprint(f'Failed to load image {f.name}: {e}') + logger.error(f'Failed to load image {f.name}: {e}') continue # Populate simple fields @@ -113,7 +131,7 @@ def load_images(input_dir: str, config: dict, load_nonlsc: bool, load_lsc: bool) continue # Handle macbeth - macbeth = locate_macbeth(config) + macbeth = locate_macbeth(image, config) if macbeth is None: continue @@ -123,3 +141,46 @@ def load_images(input_dir: str, config: dict, load_nonlsc: bool, load_lsc: bool) return None return images + + + +""" +Some code that will save virtual macbeth charts that show the difference between optimised matrices and non optimised matrices + +The function creates an image that is 1550 by 1050 pixels wide, and fills it with patches which are 200x200 pixels in size +Each patch contains the ideal color, the color from the original matrix, and the color from the final matrix +_________________ +| | +| Ideal Color | +|_______________| +| Old | new | +| Color | Color | +|_______|_______| + +Nice way of showing how the optimisation helps change the colors and the color matricies +""" +def visualise_macbeth_chart(macbeth_rgb, original_rgb, new_rgb, output_filename): + image = np.zeros((1050, 1550, 3), dtype=np.uint8) + colorindex = -1 + for y in range(6): + for x in range(4): # Creates 6 x 4 grid of macbeth chart + colorindex += 1 + xlocation = 50 + 250 * x # Means there is 50px of black gap between each square, more like the real macbeth chart. + ylocation = 50 + 250 * y + for g in range(200): + for i in range(100): + image[xlocation + i, ylocation + g] = macbeth_rgb[colorindex] + xlocation = 150 + 250 * x + ylocation = 50 + 250 * y + for i in range(100): + for g in range(100): + image[xlocation + i, ylocation + g] = original_rgb[colorindex] # Smaller squares below to compare the old colors with the new ones + xlocation = 150 + 250 * x + ylocation = 150 + 250 * y + for i in range(100): + for g in range(100): + image[xlocation + i, ylocation + g] = new_rgb[colorindex] + + im_bgr = cv2.cvtColor(image, cv2.COLOR_RGB2BGR) + cv2.imwrite(f'{output_filename} Generated Macbeth Chart.png', im_bgr) + diff --git a/utils/tuning/requirements.txt b/utils/tuning/requirements.txt new file mode 100644 index 00000000..3705769b --- /dev/null +++ b/utils/tuning/requirements.txt @@ -0,0 +1,9 @@ +coloredlogs +matplotlib +numpy +opencv-python +py3exiv2 +pyyaml +rawpy +scikit-learn +scipy diff --git a/utils/tuning/rkisp1.py b/utils/tuning/rkisp1.py index d0ce15d5..f5c42a61 100755 --- a/utils/tuning/rkisp1.py +++ b/utils/tuning/rkisp1.py @@ -5,6 +5,8 @@ # # Tuning script for rkisp1 +import coloredlogs +import logging import sys import libtuning as lt @@ -12,31 +14,42 @@ from libtuning.parsers import YamlParser from libtuning.generators import YamlOutput from libtuning.modules.lsc import LSCRkISP1 from libtuning.modules.agc import AGCRkISP1 +from libtuning.modules.ccm import CCMRkISP1 +from libtuning.modules.static import StaticModule + +coloredlogs.install(level=logging.INFO, fmt='%(name)s %(levelname)s %(message)s') + +agc = AGCRkISP1(debug=[lt.Debug.Plot]) +awb = StaticModule('Awb') +blc = StaticModule('BlackLevelCorrection') +ccm = CCMRkISP1(debug=[lt.Debug.Plot]) +color_processing = StaticModule('ColorProcessing') +filter = StaticModule('Filter') +gamma_out = StaticModule('GammaOutCorrection', {'gamma': 2.2}) +lsc = LSCRkISP1(debug=[lt.Debug.Plot], + # This is for the actual LSC tuning, and is part of the base LSC + # module. rkisp1's table sector sizes (16x16 programmed as mirrored + # 8x8) are separate, and is hardcoded in its specific LSC tuning + # module. + sector_shape=(17, 17), + + sector_x_gradient=lt.gradient.Linear(lt.Remainder.DistributeFront), + sector_y_gradient=lt.gradient.Linear(lt.Remainder.DistributeFront), + + # This is the function that will be used to average the pixels in + # each sector. This can also be a custom function. + sector_average_function=lt.average.Mean(), + + # This is the function that will be used to smooth the color ratio + # values. This can also be a custom function. + smoothing_function=lt.smoothing.MedianBlur(3),) tuner = lt.Tuner('RkISP1') -tuner.add(LSCRkISP1( - debug=[lt.Debug.Plot], - # This is for the actual LSC tuning, and is part of the base LSC - # module. rkisp1's table sector sizes (16x16 programmed as mirrored - # 8x8) are separate, and is hardcoded in its specific LSC tuning - # module. - sector_shape=(17, 17), - - sector_x_gradient=lt.gradient.Linear(lt.Remainder.DistributeFront), - sector_y_gradient=lt.gradient.Linear(lt.Remainder.DistributeFront), - - # This is the function that will be used to average the pixels in - # each sector. This can also be a custom function. - sector_average_function=lt.average.Mean(), - - # This is the function that will be used to smooth the color ratio - # values. This can also be a custom function. - smoothing_function=lt.smoothing.MedianBlur(3), - )) -tuner.add(AGCRkISP1(debug=[lt.Debug.Plot])) +tuner.add([agc, awb, blc, ccm, color_processing, filter, gamma_out, lsc]) tuner.set_input_parser(YamlParser()) tuner.set_output_formatter(YamlOutput()) -tuner.set_output_order([AGCRkISP1, LSCRkISP1]) +tuner.set_output_order([agc, awb, blc, ccm, color_processing, + filter, gamma_out, lsc]) if __name__ == '__main__': sys.exit(tuner.run(sys.argv)) |