opencv/opencv_contrib
1
0
Fork 0
mirror of https://github.com/opencv/opencv_contrib.git synced 2025-10-19 11:21:39 +08:00
Code Issues Releases Wiki Activity

Allows structured_light pipeline to be run from Python

Browse Source 
SinusoidalPattern::unwrapPhaseMap now takes an InputArray instead of InputArrayOfArrays to correct a Python binding problem
present a scriptable HistogramPhaseUnwrapping::create

replicate C++ structured_light test in Python

PhaseUnwrapping now init unwrappedPhase so pixel outside the mask area are set to 0

python binding for HistogramPhaseUnwrapping::Params to use HistogramPhaseUnwrapping::create
This commit is contained in:
David Geldreich
2020-03-08 23:18:20 +01:00
parent 9c0ae273fd
commit e967557e17
5 changed files with 109 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

94
modules/structured_light/misc/python/test/test_structured_light.py Normal file
View File

@@ -0,0 +1,94 @@
#!/usr/bin/env python
# Python 2/3 compatibility
from __future__ import print_function
import os, numpy
import cv2 as cv
from tests_common import NewOpenCVTests
class structured_light_test(NewOpenCVTests):
def test_unwrap(self):
paramsPsp = cv.structured_light_SinusoidalPattern_Params();
paramsFtp = cv.structured_light_SinusoidalPattern_Params();
paramsFaps = cv.structured_light_SinusoidalPattern_Params();
paramsPsp.methodId = cv.structured_light.PSP;
paramsFtp.methodId = cv.structured_light.FTP;
paramsFaps.methodId = cv.structured_light.FAPS;
sinusPsp = cv.structured_light.SinusoidalPattern_create(paramsPsp)
sinusFtp = cv.structured_light.SinusoidalPattern_create(paramsFtp)
sinusFaps = cv.structured_light.SinusoidalPattern_create(paramsFaps)
captures = []
for i in range(0,3):
capture = self.get_sample('/cv/structured_light/data/capture_sin_%d.jpg'%i, cv.IMREAD_GRAYSCALE)
if capture is None:
raise unittest.SkipTest("Missing files with test data")
captures.append(capture)
rows,cols = captures[0].shape
unwrappedPhaseMapPspRef = self.get_sample('/cv/structured_light/data/unwrappedPspTest.jpg',
cv.IMREAD_GRAYSCALE)
unwrappedPhaseMapFtpRef = self.get_sample('/cv/structured_light/data/unwrappedFtpTest.jpg',
cv.IMREAD_GRAYSCALE)
unwrappedPhaseMapFapsRef = self.get_sample('/cv/structured_light/data/unwrappedFapsTest.jpg',
cv.IMREAD_GRAYSCALE)
wrappedPhaseMap,shadowMask = sinusPsp.computePhaseMap(captures);
unwrappedPhaseMap = sinusPsp.unwrapPhaseMap(wrappedPhaseMap, (cols, rows), shadowMask=shadowMask)
unwrappedPhaseMap8 = unwrappedPhaseMap*1 + 128
unwrappedPhaseMap8 = numpy.uint8(unwrappedPhaseMap8)
sumOfDiff = 0
count = 0
for i in range(rows):
for j in range(cols):
ref = int(unwrappedPhaseMapPspRef[i, j])
comp = int(unwrappedPhaseMap8[i, j])
sumOfDiff += (ref - comp)
count += 1
ratio = sumOfDiff/float(count)
self.assertLessEqual(ratio, 0.2)
wrappedPhaseMap,shadowMask = sinusFtp.computePhaseMap(captures);
unwrappedPhaseMap = sinusFtp.unwrapPhaseMap(wrappedPhaseMap, (cols, rows), shadowMask=shadowMask)
unwrappedPhaseMap8 = unwrappedPhaseMap*1 + 128
unwrappedPhaseMap8 = numpy.uint8(unwrappedPhaseMap8)
sumOfDiff = 0
count = 0
for i in range(rows):
for j in range(cols):
ref = int(unwrappedPhaseMapFtpRef[i, j])
comp = int(unwrappedPhaseMap8[i, j])
sumOfDiff += (ref - comp)
count += 1
ratio = sumOfDiff/float(count)
self.assertLessEqual(ratio, 0.2)
wrappedPhaseMap,shadowMask2 = sinusFaps.computePhaseMap(captures);
unwrappedPhaseMap = sinusFaps.unwrapPhaseMap(wrappedPhaseMap, (cols, rows), shadowMask=shadowMask)
unwrappedPhaseMap8 = unwrappedPhaseMap*1 + 128
unwrappedPhaseMap8 = numpy.uint8(unwrappedPhaseMap8)
sumOfDiff = 0
count = 0
for i in range(rows):
for j in range(cols):
ref = int(unwrappedPhaseMapFapsRef[i, j])
comp = int(unwrappedPhaseMap8[i, j])
sumOfDiff += (ref - comp)
count += 1
ratio = sumOfDiff/float(count)
self.assertLessEqual(ratio, 0.2)
if __name__ == '__main__':
NewOpenCVTests.bootstrap()