brain_vision's profile - activity

I found out the recent master is not usable if one want to use those functions with python located in nonfree/contrib. Because most of those functions are blacklisted since 11/08/2014. Hope they will come back soon.

in modules/python/common.cmake

# module blacklist

ocv_list_filterout(candidate_deps "^opencv_cud(a|ev)")
ocv_list_filterout(candidate_deps "^opencv_adas$")
ocv_list_filterout(candidate_deps "^opencv_face$")
ocv_list_filterout(candidate_deps "^opencv_matlab$")
ocv_list_filterout(candidate_deps "^opencv_tracking$")
ocv_list_filterout(candidate_deps "^opencv_optflow$")
ocv_list_filterout(candidate_deps "^opencv_bgsegm$")
ocv_list_filterout(candidate_deps "^opencv_xfeatures2d$")
ocv_list_filterout(candidate_deps "^opencv_ximgproc$")
ocv_list_filterout(candidate_deps "^opencv_xphoto$")
ocv_list_filterout(candidate_deps "^opencv_ts$")

ocv_add_module(${MODULE_NAME} BINDINGS OPTIONAL ${candidate_deps})

ocv_module_include_directories(
    "${PYTHON_INCLUDE_PATH}"
    ${PYTHON_NUMPY_INCLUDE_DIRS}
    "${PYTHON_SOURCE_DIR}/src2"
    )

@berak Thanks!

I cannot build the nonfree modules in opencv 3.0. The -DBUILD_opencv_nonfree:BOOL="1" has no effect at all. How can I build them?

Are they expected, or there is some configuration in my system to cause those errors?

failed with 1 extra row:

failed with Argument "xmap" has unexpected number of rows:

OCL_StitchingWarpersFixture_StitchingWarpers_BuildMaps.StitchingWarpers_BuildMaps/1

(639 instead of 638)

OCL_StitchingWarpersFixture_StitchingWarpers_BuildMaps.StitchingWarpers_BuildMaps/4, /7, /10

OCL_StitchingWarpersFixture_StitchingWarpers_Warp.StitchingWarpers_Warp/1, /4, /7, /10 with Argument "dst" has unexpected number of rows

failed with unexpected type:

Argument "dst" has unexpected type 16 instead of 24

Size_CvtMode2_cvtColorYUV420.cvtColorYUV420/19, /20, /25, /26, /27, /28, /48, /49, /54--/57, /77, /78, /83--/86

compiled with gcc 4.8.3

-DENABLE_FAST_MATH:BOOL="1"

-DENABLE_AVX:BOOL="1"

-DENABLE_COVERAGE:BOOL="1"

-DWITH_OPENGL:BOOL="1"

-DWITH_QT:BOOL="1"

-DWITH_TBB:BOOL="1"

test command:

cd build python

../modules/ts/misc/run.py --check

Are they expected, or there is some configuration in my system to cause those errors?

1. failed with:

   Bad flag (parameter or structure field) (Unrecognized or unsupported array type) in cvGetMat

   Shape_SCD.regression

   Hauss.regression

   ShapeEMD_SCD.regression

2. write_imageseq has Unexpected output: Imgcodecs_Image.write_imageseq

   [ RUN ]

   Imgcodecs_Image.write_imageseq

   opencv/modules/ts/src/ts.cpp:515:

   Failure Failed

   failure reason: Unexpected output test case #-1 seed: ffffffffffffffff ----------------------------------- LOG:

   ... #all failed

   ---

   [ FAILED ] Imgcodecs_Image.write_imageseq (5377 ms)

3. failed with wrong accuracy:

[ RUN ]

Video_calcOpticalFlowDual_TVL1.Regression

opencv/modules/video/test/test_tvl1optflow.cpp:170:

Failure Expected: (err) <= (MAX_RMSE),

actual: 0.0309178 vs 0.03 [ FAILED ]

Video_calcOpticalFlowDual_TVL1.Regression

(1541 ms)

compiled with gcc 4.8.3

-DENABLE_FAST_MATH:BOOL="1"

-DENABLE_AVX:BOOL="1"

-DENABLE_COVERAGE:BOOL="1"

-DWITH_OPENGL:BOOL="1"

-DWITH_QT:BOOL="1"

-DWITH_TBB:BOOL="1"

test command:

cd build python

../modules/ts/misc/run.py -a

Can anyone confirm this? In 2.4.9 lots of functions are missing for example: cv2.drawMatches

Here is the fail message:

###########
#opencv_test_video
###########
[ RUN      ] Video_calcOpticalFlowDual_TVL1.Regression
/home/dracula/mytmp/opencv/modules/video/test/test_tvl1optflow.cpp:169: Failure
Expected: (err) <= (MAX_RMSE), actual: 0.0207956 vs 0.02
[  FAILED  ] Video_calcOpticalFlowDual_TVL1.Regression (1069 ms)

compiled with gcc 4.8.3

-DENABLE_FAST_MATH:BOOL="1" 
-DENABLE_AVX:BOOL="1" 
-DENABLE_COVERAGE:BOOL="1"
-DWITH_OPENGL:BOOL="1" 
-DWITH_QT:BOOL="1" 
-DWITH_TBB:BOOL="1"

versions:

opencv: f8758da289c092fcce10baa4723d6388c4e41cb3 2.4.9.x-prep

opencv_extra: 61d59fcaa8d3cebb05ae6eadf0ab8cc5ac597445 tags/2.4.9

Can anyone tell me if this is because I used wrong opencv_extra version? Which version I should use?