1# Copyright 2013 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 math
16import os.path
17import its.caps
18import its.device
19import its.image
20import its.objects
21import its.target
22import matplotlib
23from matplotlib import pylab
24import numpy
25
26NAME = os.path.basename(__file__).split('.')[0]
27RESIDUAL_THRESHOLD = 0.0003  # approximately each sample is off by 2/255
28# The HAL3.2 spec requires that curves up to 64 control points in length
29# must be supported.
30L = 64
31LM1 = float(L-1)
32
33
34def main():
35    """Test that device processing can be inverted to linear pixels.
36
37    Captures a sequence of shots with the device pointed at a uniform
38    target. Attempts to invert all the ISP processing to get back to
39    linear R,G,B pixel data.
40    """
41    gamma_lut = numpy.array(
42            sum([[i/LM1, math.pow(i/LM1, 1/2.2)] for i in xrange(L)], []))
43    inv_gamma_lut = numpy.array(
44            sum([[i/LM1, math.pow(i/LM1, 2.2)] for i in xrange(L)], []))
45
46    with its.device.ItsSession() as cam:
47        props = cam.get_camera_properties()
48        props = cam.override_with_hidden_physical_camera_props(props)
49        its.caps.skip_unless(its.caps.compute_target_exposure(props))
50        sync_latency = its.caps.sync_latency(props)
51
52        debug = its.caps.debug_mode()
53        largest_yuv = its.objects.get_largest_yuv_format(props)
54        if debug:
55            fmt = largest_yuv
56        else:
57            match_ar = (largest_yuv['width'], largest_yuv['height'])
58            fmt = its.objects.get_smallest_yuv_format(props, match_ar=match_ar)
59
60        e, s = its.target.get_target_exposure_combos(cam)['midSensitivity']
61        s /= 2
62        sens_range = props['android.sensor.info.sensitivityRange']
63        sensitivities = [s*1.0/3.0, s*2.0/3.0, s, s*4.0/3.0, s*5.0/3.0]
64        sensitivities = [s for s in sensitivities
65                         if s > sens_range[0] and s < sens_range[1]]
66
67        req = its.objects.manual_capture_request(0, e)
68        req['android.blackLevel.lock'] = True
69        req['android.tonemap.mode'] = 0
70        req['android.tonemap.curve'] = {
71                'red': gamma_lut.tolist(),
72                'green': gamma_lut.tolist(),
73                'blue': gamma_lut.tolist()}
74
75        r_means = []
76        g_means = []
77        b_means = []
78
79        for sens in sensitivities:
80            req['android.sensor.sensitivity'] = sens
81            cap = its.device.do_capture_with_latency(
82                    cam, req, sync_latency, fmt)
83            img = its.image.convert_capture_to_rgb_image(cap)
84            its.image.write_image(
85                    img, '%s_sens=%04d.jpg' % (NAME, sens))
86            img = its.image.apply_lut_to_image(img, inv_gamma_lut[1::2] * LM1)
87            tile = its.image.get_image_patch(img, 0.45, 0.45, 0.1, 0.1)
88            rgb_means = its.image.compute_image_means(tile)
89            r_means.append(rgb_means[0])
90            g_means.append(rgb_means[1])
91            b_means.append(rgb_means[2])
92
93        pylab.title(NAME)
94        pylab.plot(sensitivities, r_means, '-ro')
95        pylab.plot(sensitivities, g_means, '-go')
96        pylab.plot(sensitivities, b_means, '-bo')
97        pylab.xlim([sens_range[0], sens_range[1]/2])
98        pylab.ylim([0, 1])
99        pylab.xlabel('sensitivity(ISO)')
100        pylab.ylabel('RGB avg [0, 1]')
101        matplotlib.pyplot.savefig('%s_plot_means.png' % (NAME))
102
103        # Check that each plot is actually linear.
104        for means in [r_means, g_means, b_means]:
105            line, residuals, _, _, _ = numpy.polyfit(range(len(sensitivities)),
106                                                     means, 1, full=True)
107            print 'Line: m=%f, b=%f, resid=%f'%(line[0], line[1], residuals[0])
108            msg = 'residual: %.5f, THRESH: %.4f' % (
109                    residuals[0], RESIDUAL_THRESHOLD)
110            assert residuals[0] < RESIDUAL_THRESHOLD, msg
111
112if __name__ == '__main__':
113    main()
114
115