# Copyright 2013 The Android Open Source Project # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. """Verifies YUV & JPEG image captures have similar brightness.""" import logging import os.path import matplotlib from matplotlib import pylab import matplotlib.lines as mlines from matplotlib.ticker import MaxNLocator from mobly import test_runner import its_base_test import camera_properties_utils import capture_request_utils import image_processing_utils import its_session_utils import target_exposure_utils _JPG_STR = 'jpg' _NAME = os.path.splitext(os.path.basename(__file__))[0] _PATCH_H = 0.1 # center 10% _PATCH_W = 0.1 _PATCH_X = 0.5 - _PATCH_W/2 _PATCH_Y = 0.5 - _PATCH_H/2 _PLOT_ALPHA = 0.5 _PLOT_MARKER_SIZE = 8 _PLOT_LEGEND_CIRCLE_SIZE = 10 _PLOT_LEGEND_TRIANGLE_SIZE = 6 _THRESHOLD_MAX_RMS_DIFF = 0.03 _YUV_STR = 'yuv' def do_capture_and_extract_rgb_means( req, cam, props, size, img_type, index, name_with_log_path, debug): """Do capture and extra rgb_means of center patch. Args: req: capture request cam: camera object props: camera properties dict size: [width, height] img_type: string of 'yuv' or 'jpeg' index: index to track capture number of img_type name_with_log_path: file name and location for saving image debug: boolean to flag saving captured images Returns: rgb: center patch RGB means img: RGB image array """ out_surface = {'width': size[0], 'height': size[1], 'format': img_type} if camera_properties_utils.stream_use_case(props): out_surface['useCase'] = camera_properties_utils.USE_CASE_STILL_CAPTURE logging.debug('output surface: %s', str(out_surface)) if debug and camera_properties_utils.raw(props): out_surfaces = [{'format': 'raw'}, out_surface] cap_raw, cap = cam.do_capture(req, out_surfaces) img_raw = image_processing_utils.convert_capture_to_rgb_image( cap_raw, props=props) image_processing_utils.write_image( img_raw, f'{name_with_log_path}_raw_{img_type}_w{size[0]}_h{size[1]}.png', True) else: cap = cam.do_capture(req, out_surface) logging.debug('e_cap: %d, s_cap: %d, f_distance: %s', cap['metadata']['android.sensor.exposureTime'], cap['metadata']['android.sensor.sensitivity'], cap['metadata']['android.lens.focusDistance']) if img_type == _JPG_STR: if cap['format'] != 'jpeg': raise AssertionError(f"{cap['format']} != jpeg") img = image_processing_utils.decompress_jpeg_to_rgb_image(cap['data']) else: if cap['format'] != img_type: raise AssertionError(f"{cap['format']} != {img_type}") img = image_processing_utils.convert_capture_to_rgb_image(cap) if cap['width'] != size[0]: raise AssertionError(f"{cap['width']} != {size[0]}") if cap['height'] != size[1]: raise AssertionError(f"{cap['height']} != {size[1]}") if img_type == _JPG_STR: if img.shape[0] != size[1]: raise AssertionError(f'{img.shape[0]} != {size[1]}') if img.shape[1] != size[0]: raise AssertionError(f'{img.shape[1]} != {size[0]}') if img.shape[2] != 3: raise AssertionError(f'{img.shape[2]} != 3') patch = image_processing_utils.get_image_patch( img, _PATCH_X, _PATCH_Y, _PATCH_W, _PATCH_H) rgb = image_processing_utils.compute_image_means(patch) logging.debug('Captured %s %dx%d rgb = %s, format number = %d', img_type, cap['width'], cap['height'], str(rgb), index) return rgb, img class YuvJpegAllTest(its_base_test.ItsBaseTest): """Test reported sizes & fmts for YUV & JPEG caps return similar images.""" def test_yuv_jpeg_all(self): with its_session_utils.ItsSession( device_id=self.dut.serial, camera_id=self.camera_id, hidden_physical_id=self.hidden_physical_id) as cam: props = cam.get_camera_properties() props = cam.override_with_hidden_physical_camera_props(props) log_path = self.log_path debug = self.debug_mode name_with_log_path = os.path.join(log_path, _NAME) # Check SKIP conditions camera_properties_utils.skip_unless( camera_properties_utils.linear_tonemap(props)) # Load chart for scene its_session_utils.load_scene( cam, props, self.scene, self.tablet, its_session_utils.CHART_DISTANCE_NO_SCALING) # If device supports target exposure computation, use manual capture. # Otherwise, do 3A, then use an auto request. # Both requests use a linear tonemap and focus distance of 0.0 # so that the YUV and JPEG should look the same # (once converted by the image_processing_utils). if camera_properties_utils.compute_target_exposure(props): logging.debug('Using manual capture request') e, s = target_exposure_utils.get_target_exposure_combos( log_path, cam)['midExposureTime'] logging.debug('e_req: %d, s_req: %d', e, s) req = capture_request_utils.manual_capture_request( s, e, 0.0, True, props) match_ar = None else: logging.debug('Using auto capture request') cam.do_3a(do_af=False) req = capture_request_utils.auto_capture_request( linear_tonemap=True, props=props, do_af=False) largest_yuv = capture_request_utils.get_largest_yuv_format(props) match_ar = (largest_yuv['width'], largest_yuv['height']) yuv_rgbs = [] yuv_imgs = [] for i, size in enumerate( capture_request_utils.get_available_output_sizes( _YUV_STR, props, match_ar_size=match_ar)): yuv_rgb, yuv_img = do_capture_and_extract_rgb_means( req, cam, props, size, _YUV_STR, i, name_with_log_path, debug) yuv_rgbs.append(yuv_rgb) yuv_imgs.append(yuv_img) jpg_rgbs = [] jpg_imgs = [] for i, size in enumerate( capture_request_utils.get_available_output_sizes( _JPG_STR, props, match_ar_size=match_ar)): jpg_rgb, jpg_img = do_capture_and_extract_rgb_means( req, cam, props, size, _JPG_STR, i, name_with_log_path, debug) jpg_rgbs.append(jpg_rgb) jpg_imgs.append(jpg_img) # Plot means vs format pylab.figure(_NAME) pylab.title(_NAME) yuv_index = range(len(yuv_rgbs)) jpg_index = range(len(jpg_rgbs)) pylab.plot(yuv_index, [rgb[0] for rgb in yuv_rgbs], '-ro', alpha=_PLOT_ALPHA, markersize=_PLOT_MARKER_SIZE) pylab.plot(yuv_index, [rgb[1] for rgb in yuv_rgbs], '-go', alpha=_PLOT_ALPHA, markersize=_PLOT_MARKER_SIZE) pylab.plot(yuv_index, [rgb[2] for rgb in yuv_rgbs], '-bo', alpha=_PLOT_ALPHA, markersize=_PLOT_MARKER_SIZE) pylab.plot(jpg_index, [rgb[0] for rgb in jpg_rgbs], '-r^', alpha=_PLOT_ALPHA, markersize=_PLOT_MARKER_SIZE) pylab.plot(jpg_index, [rgb[1] for rgb in jpg_rgbs], '-g^', alpha=_PLOT_ALPHA, markersize=_PLOT_MARKER_SIZE) pylab.plot(jpg_index, [rgb[2] for rgb in jpg_rgbs], '-b^', alpha=_PLOT_ALPHA, markersize=_PLOT_MARKER_SIZE) pylab.ylim([0, 1]) ax = pylab.gca() ax.xaxis.set_major_locator(MaxNLocator(integer=True)) # x-axis integers yuv_marker = mlines.Line2D([], [], linestyle='None', color='black', marker='.', markersize=_PLOT_LEGEND_CIRCLE_SIZE, label='YUV') jpg_marker = mlines.Line2D([], [], linestyle='None', color='black', marker='^', markersize=_PLOT_LEGEND_TRIANGLE_SIZE, label='JPEG') ax.legend(handles=[yuv_marker, jpg_marker]) pylab.xlabel('format number') pylab.ylabel('RGB avg [0, 1]') matplotlib.pyplot.savefig(f'{name_with_log_path}_plot_means.png') # Assert all captures are similar in RGB space using rgbs[0] as ref. rgbs = yuv_rgbs + jpg_rgbs max_diff = 0 for rgb_i in rgbs[1:]: rms_diff = image_processing_utils.compute_image_rms_difference_1d( rgbs[0], rgb_i) # use first capture as reference max_diff = max(max_diff, rms_diff) msg = f'Max RMS difference: {max_diff:.4f}' logging.debug('%s', msg) if max_diff >= _THRESHOLD_MAX_RMS_DIFF: for img in yuv_imgs: image_processing_utils.write_image( img, f'{name_with_log_path}_yuv_{img.shape[1]}x{img.shape[0]}.png' ) for img in jpg_imgs: image_processing_utils.write_image( img, f'{name_with_log_path}_jpg_{img.shape[1]}x{img.shape[0]}.png' ) raise AssertionError(f'{msg} spec: {_THRESHOLD_MAX_RMS_DIFF}') if __name__ == '__main__': test_runner.main()