1# Copyright 2014 The Android Open Source Project 2# 3# Licensed under the Apache License, Version 2.0 (the "License"); 4# you may not use this file except in compliance with the License. 5# You may obtain a copy of the License at 6# 7# http://www.apache.org/licenses/LICENSE-2.0 8# 9# Unless required by applicable law or agreed to in writing, software 10# distributed under the License is distributed on an "AS IS" BASIS, 11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 12# See the License for the specific language governing permissions and 13# limitations under the License. 14 15import its.image 16import its.caps 17import its.device 18import its.objects 19import os.path 20import numpy 21 22def main(): 23 """Take long bursts of images and check that they're all identical. 24 25 Assumes a static scene. Can be used to idenfity if there are sporadic 26 frames that are processed differently or have artifacts, or if 3A isn't 27 stable, since this test converges 3A at the start but doesn't lock 3A 28 throughout capture. 29 """ 30 NAME = os.path.basename(__file__).split(".")[0] 31 32 BURST_LEN = 50 33 BURSTS = 5 34 FRAMES = BURST_LEN * BURSTS 35 36 SPREAD_THRESH = 0.03 37 38 with its.device.ItsSession() as cam: 39 40 # Capture at the smallest resolution. 41 props = cam.get_camera_properties() 42 its.caps.skip_unless(its.caps.manual_sensor(props) and 43 its.caps.awb_lock(props)) 44 45 _, fmt = its.objects.get_fastest_manual_capture_settings(props) 46 w,h = fmt["width"], fmt["height"] 47 48 # Converge 3A prior to capture. 49 cam.do_3a(lock_ae=True, lock_awb=True) 50 51 # After 3A has converged, lock AE+AWB for the duration of the test. 52 req = its.objects.fastest_auto_capture_request(props) 53 req["android.blackLevel.lock"] = True 54 req["android.control.awbLock"] = True 55 req["android.control.aeLock"] = True 56 57 # Capture bursts of YUV shots. 58 # Get the mean values of a center patch for each. 59 # Also build a 4D array, which is an array of all RGB images. 60 r_means = [] 61 g_means = [] 62 b_means = [] 63 imgs = numpy.empty([FRAMES,h,w,3]) 64 for j in range(BURSTS): 65 caps = cam.do_capture([req]*BURST_LEN, [fmt]) 66 for i,cap in enumerate(caps): 67 n = j*BURST_LEN + i 68 imgs[n] = its.image.convert_capture_to_rgb_image(cap) 69 tile = its.image.get_image_patch(imgs[n], 0.45, 0.45, 0.1, 0.1) 70 means = its.image.compute_image_means(tile) 71 r_means.append(means[0]) 72 g_means.append(means[1]) 73 b_means.append(means[2]) 74 75 # Dump all images. 76 print "Dumping images" 77 for i in range(FRAMES): 78 its.image.write_image(imgs[i], "%s_frame%03d.jpg"%(NAME,i)) 79 80 # The mean image. 81 img_mean = imgs.mean(0) 82 its.image.write_image(img_mean, "%s_mean.jpg"%(NAME)) 83 84 # Pass/fail based on center patch similarity. 85 for means in [r_means, g_means, b_means]: 86 spread = max(means) - min(means) 87 print spread 88 assert(spread < SPREAD_THRESH) 89 90if __name__ == '__main__': 91 main() 92 93