求助！BUILDROOT 编译python3 OPENCV3到ARM上，imread出错【100元悬赏解决方案，有效期至12月9日】

k455619 · 2020-11-26 10:45:03

raspberryman · 2020-11-26 11:10:07

全勾上吧

k455619 · 2020-11-26 11:18:17

raspberryman 说：

全勾上吧

全勾上了，正在编译，感觉没得啥子用，但是可以一试

k455619 · 2020-11-26 13:36:58

nand flash 太小了这样并不能

raspberryman · 2020-11-26 15:05:57

在 ubuntu@x86 用 chroot 到 ARM 平台的 rootfs
http://whycan.com/t_3251.html
(出处:哇酷开发者社区【VMWare/Linux/Ubuntu/Fedora/CentOS/U-BOOT】)

可以先用Ubuntu直接测试

raspberryman · 2020-11-26 15:18:10

I tried so many hacks. They were not working. Someone suggested:

from cv2 import cv2
I think this is the best solution to this problem

https://stackoverflow.com/questions/51593147/cv2-python-has-no-imread-member

k455619 · 2020-11-26 15:25:03

raspberryman 说：

I tried so many hacks. They were not working. Someone suggested:
from cv2 import cv2
I think this is the best solution to this problem

https://stackoverflow.com/questions/51593147/cv2-python-has-no-imread-member

在ARM板上执行:
from cv2 import cv2
random: python: uninitialized urandom read (4096 bytes read)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ImportError: cannot import name 'cv2' from 'cv2' (/usr/lib/python3.8/site-packages/cv2/python-3.8/cv2.cpython-38-arm-linux-gnueabihf.so)

k455619 · 2020-11-26 15:28:07

事实上ARM板上根本没有im相关的方法,我不知道是不是我哪里配置错了。

>>> print(dir(cv2))
['', 'ACCESS_FAST', 'ACCESS_MASK', 'ACCESS_READ', 'ACCESS_RW', 'ACCESS_WRITE', 'Algorithm', 'AsyncArray', 'BORDER_CONSTANT', 'BORDER_DEFAULT', 'BORDER_ISOLATED', 'BORDER_REFLECT', 'BORDER_REFLECT101', 'BORDER_REFLECT_101', 'BORDER_REPLICATE', 'BORDER_TRANSPARENT', 'BORDER_WRAP', 'CMP_EQ', 'CMP_GE', 'CMP_GT', 'CMP_LE', 'CMP_LT', 'CMP_NE', 'COVAR_COLS', 'COVAR_NORMAL', 'COVAR_ROWS', 'COVAR_SCALE', 'COVAR_SCRAMBLED', 'COVAR_USE_AVG', 'CV_16S', 'CV_16SC1', 'CV_16SC2', 'CV_16SC3', 'CV_16SC4', 'CV_16U', 'CV_16UC1', 'CV_16UC2', 'CV_16UC3', 'CV_16UC4', 'CV_32F', 'CV_32FC1', 'CV_32FC2', 'CV_32FC3', 'CV_32FC4', 'CV_32S', 'CV_32SC1', 'CV_32SC2', 'CV_32SC3', 'CV_32SC4', 'CV_64F', 'CV_64FC1', 'CV_64FC2', 'CV_64FC3', 'CV_64FC4', 'CV_8S', 'CV_8SC1', 'CV_8SC2', 'CV_8SC3', 'CV_8SC4', 'CV_8U', 'CV_8UC1', 'CV_8UC2', 'CV_8UC3', 'CV_8UC4', 'DCT_INVERSE', 'DCT_ROWS', 'DECOMP_CHOLESKY', 'DECOMP_EIG', 'DECOMP_LU', 'DECOMP_NORMAL', 'DECOMP_QR', 'DECOMP_SVD', 'DFT_COMPLEX_INPUT', 'DFT_COMPLEX_OUTPUT', 'DFT_INVERSE', 'DFT_REAL_OUTPUT', 'DFT_ROWS', 'DFT_SCALE', 'DMatch', 'Error', 'FILE_NODE_EMPTY', 'FILE_NODE_FLOAT', 'FILE_NODE_FLOW', 'FILE_NODE_INT', 'FILE_NODE_MAP', 'FILE_NODE_NAMED', 'FILE_NODE_NONE', 'FILE_NODE_REAL', 'FILE_NODE_REF', 'FILE_NODE_SEQ', 'FILE_NODE_STR', 'FILE_NODE_STRING', 'FILE_NODE_TYPE_MASK', 'FILE_NODE_USER', 'FILE_STORAGE_APPEND', 'FILE_STORAGE_BASE64', 'FILrandom: fast init done
E_STORAGE_FORMAT_AUTO', 'FILE_STORAGE_FORMAT_JSON', 'FILE_STORAGE_FORMAT_MASK', 'FILE_STORAGE_FORMAT_XML', 'FILE_STORAGE_FORMAT_YAML', 'FILE_STORAGE_INSIDE_MAP', 'FILE_STORAGE_MEMORY', 'FILE_STORAGE_NAME_EXPECTED', 'FILE_STORAGE_READ', 'FILE_STORAGE_UNDEFINED', 'FILE_STORAGE_VALUE_EXPECTED', 'FILE_STORAGE_WRITE', 'FILE_STORAGE_WRITE_BASE64', 'FILLED', 'FONT_HERSHEY_COMPLEX', 'FONT_HERSHEY_COMPLEX_SMALL', 'FONT_HERSHEY_DUPLEX', 'FONT_HERSHEY_PLAIN', 'FONT_HERSHEY_SCRIPT_COMPLEX', 'FONT_HERSHEY_SCRIPT_SIMPLEX', 'FONT_HERSHEY_SIMPLEX', 'FONT_HERSHEY_TRIPLEX', 'FONT_ITALIC', 'FORMATTER_FMT_C', 'FORMATTER_FMT_CSV', 'FORMATTER_FMT_DEFAULT', 'FORMATTER_FMT_MATLAB', 'FORMATTER_FMT_NUMPY', 'FORMATTER_FMT_PYTHON', 'FileNode', 'FileNode_EMPTY', 'FileNode_FLOAT', 'FileNode_FLOW', 'FileNode_INT', 'FileNode_MAP', 'FileNode_NAMED', 'FileNode_NONE', 'FileNode_REAL', 'FileNode_REF', 'FileNode_SEQ', 'FileNode_STR', 'FileNode_STRING', 'FileNode_TYPE_MASK', 'FileNode_USER', 'FileStorage', 'FileStorage_APPEND', 'FileStorage_BASE64', 'FileStorage_FORMAT_AUTO', 'FileStorage_FORMAT_JSON', 'FileStorage_FORMAT_MASK', 'FileStorage_FORMAT_XML', 'FileStorage_FORMAT_YAML', 'FileStorage_INSIDE_MAP', 'FileStorage_MEMORY', 'FileStorage_NAME_EXPECTED', 'FileStorage_READ', 'FileStorage_UNDEFINED', 'FileStorage_VALUE_EXPECTED', 'FileStorage_WRITE', 'FileStorage_WRITE_BASE64', 'Formatter_FMT_C', 'Formatter_FMT_CSV', 'Formatter_FMT_DEFAULT', 'Formatter_FMT_MATLAB', 'Formatter_FMT_NUMPY', 'Formatter_FMT_PYTHON', 'GEMM_1_T', 'GEMM_2_T', 'GEMM_3_T', 'HAMMING_NORM_TYPE', 'Hamming_normType', 'KMEANS_PP_CENTERS', 'KMEANS_RANDOM_CENTERS', 'KMEANS_USE_INITIAL_LABELS', 'KeyPoint', 'KeyPoint_convert', 'KeyPoint_overlap', 'LINE_4', 'LINE_8', 'LINE_AA', 'LUT', 'MAT_AUTO_STEP', 'MAT_CONTINUOUS_FLAG', 'MAT_DEPTH_MASK', 'MAT_MAGIC_MASK', 'MAT_MAGIC_VAL', 'MAT_SUBMATRIX_FLAG', 'MAT_TYPE_MASK', 'Mahalanobis', 'Mat_AUTO_STEP', 'Mat_CONTINUOUS_FLAG', 'Mat_DEPTH_MASK', 'Mat_MAGIC_MASK', 'Mat_MAGIC_VAL', 'Mat_SUBMATRIX_FLAG', 'Mat_TYPE_MASK', 'NORM_HAMMING', 'NORM_HAMMING2', 'NORM_INF', 'NORM_L1', 'NORM_L2', 'NORM_L2SQR', 'NORM_MINMAX', 'NORM_RELATIVE', 'NORM_TYPE_MASK', 'PARAM_ALGORITHM', 'PARAM_BOOLEAN', 'PARAM_FLOAT', 'PARAM_INT', 'PARAM_MAT', 'PARAM_MAT_VECTOR', 'PARAM_REAL', 'PARAM_SCALAR', 'PARAM_STRING', 'PARAM_UCHAR', 'PARAM_UINT64', 'PARAM_UNSIGNED_INT', 'PCABackProject', 'PCACompute', 'PCACompute2', 'PCAProject', 'PCA_DATA_AS_COL', 'PCA_DATA_AS_ROW', 'PCA_USE_AVG', 'PSNR', 'Param_ALGORITHM', 'Param_BOOLEAN', 'Param_FLOAT', 'Param_INT', 'Param_MAT', 'Param_MAT_VECTOR', 'Param_REAL', 'Param_SCALAR', 'Param_STRING', 'Param_UCHAR', 'Param_UINT64', 'Param_UNSIGNED_INT', 'REDUCE_AVG', 'REDUCE_MAX', 'REDUCE_MIN', 'REDUCE_SUM', 'RNG_NORMAL', 'RNG_UNIFORM', 'ROTATE_180', 'ROTATE_90_CLOCKWISE', 'ROTATE_90_COUNTERCLOCKWISE', 'SOLVELP_MULTI', 'SOLVELP_SINGLE', 'SOLVELP_UNBOUNDED', 'SOLVELP_UNFEASIBLE', 'SORT_ASCENDING', 'SORT_DESCENDING', 'SORT_EVERY_COLUMN', 'SORT_EVERY_ROW', 'SPARSE_MAT_HASH_BIT', 'SPARSE_MAT_HASH_SCALE', 'SPARSE_MAT_MAGIC_VAL', 'SPARSE_MAT_MAX_DIM', 'SVBackSubst', 'SVD_FULL_UV', 'SVD_MODIFY_A', 'SVD_NO_UV', 'SVDecomp', 'SparseMat_HASH_BIT', 'SparseMat_HASH_SCALE', 'SparseMat_MAGIC_VAL', 'SparseMat_MAX_DIM', 'TERM_CRITERIA_COUNT', 'TERM_CRITERIA_EPS', 'TERM_CRITERIA_MAX_ITER', 'TermCriteria_COUNT', 'TermCriteria_EPS', 'TermCriteria_MAX_ITER', 'TickMeter', 'UMAT_AUTO_STEP', 'UMAT_CONTINUOUS_FLAG', 'UMAT_DATA_ASYNC_CLEANUP', 'UMAT_DATA_COPY_ON_MAP', 'UMAT_DATA_DEVICE_COPY_OBSOLETE', 'UMAT_DATA_DEVICE_MEM_MAPPED', 'UMAT_DATA_HOST_COPY_OBSOLETE', 'UMAT_DATA_TEMP_COPIED_UMAT', 'UMAT_DATA_TEMP_UMAT', 'UMAT_DATA_USER_ALLOCATED', 'UMAT_DEPTH_MASK', 'UMAT_MAGIC_MASK', 'UMAT_MAGIC_VAL', 'UMAT_SUBMATRIX_FLAG', 'UMAT_TYPE_MASK', 'UMat', 'UMatData_ASYNC_CLEANUP', 'UMatData_COPY_ON_MAP', 'UMatData_DEVICE_COPY_OBSOLETE', 'UMatData_DEVICE_MEM_MAPPED', 'UMatData_HOST_COPY_OBSOLETE', 'UMatData_TEMP_COPIED_UMAT', 'UMatData_TEMP_UMAT', 'UMatData_USER_ALLOCATED', 'UMat_AUTO_STEP', 'UMat_CONTINUOUS_FLAG', 'UMat_DEPTH_MASK', 'UMat_MAGIC_MASK', 'UMat_MAGIC_VAL', 'UMat_SUBMATRIX_FLAG', 'UMat_TYPE_MASK', 'UMat_context', 'UMat_queue', 'USAGE_ALLOCATE_DEVICE_MEMORY', 'USAGE_ALLOCATE_HOST_MEMORY', 'USAGE_ALLOCATE_SHARED_MEMORY', 'USAGE_DEFAULT', '_INPUT_ARRAY_CUDA_GPU_MAT', '_INPUT_ARRAY_CUDA_HOST_MEM', '_INPUT_ARRAY_EXPR', '_INPUT_ARRAY_FIXED_SIZE', '_INPUT_ARRAY_FIXED_TYPE', '_INPUT_ARRAY_KIND_MASK', '_INPUT_ARRAY_KIND_SHIFT', '_INPUT_ARRAY_MAT', '_INPUT_ARRAY_MATX', '_INPUT_ARRAY_NONE', '_INPUT_ARRAY_OPENGL_BUFFER', '_INPUT_ARRAY_STD_ARRAY', '_INPUT_ARRAY_STD_ARRAY_MAT', '_INPUT_ARRAY_STD_BOOL_VECTOR', '_INPUT_ARRAY_STD_VECTOR', '_INPUT_ARRAY_STD_VECTOR_CUDA_GPU_MAT', '_INPUT_ARRAY_STD_VECTOR_MAT', '_INPUT_ARRAY_STD_VECTOR_UMAT', '_INPUT_ARRAY_STD_VECTOR_VECTOR', '_INPUT_ARRAY_UMAT', '_InputArray_CUDA_GPU_MAT', '_InputArray_CUDA_HOST_MEM', '_InputArray_EXPR', '_InputArray_FIXED_SIZE', '_InputArray_FIXED_TYPE', '_InputArray_KIND_MASK', '_InputArray_KIND_SHIFT', '_InputArray_MAT', '_InputArray_MATX', '_InputArray_NONE', '_InputArray_OPENGL_BUFFER', '_InputArray_STD_ARRAY', '_InputArray_STD_ARRAY_MAT', '_InputArray_STD_BOOL_VECTOR', '_InputArray_STD_VECTOR', '_InputArray_STD_VECTOR_CUDA_GPU_MAT', '_InputArray_STD_VECTOR_MAT', '_InputArray_STD_VECTOR_UMAT', '_InputArray_STD_VECTOR_VECTOR', '_InputArray_UMAT', '_OUTPUT_ARRAY_DEPTH_MASK_16S', '_OUTPUT_ARRAY_DEPTH_MASK_16U', '_OUTPUT_ARRAY_DEPTH_MASK_32F', '_OUTPUT_ARRAY_DEPTH_MASK_32S', '_OUTPUT_ARRAY_DEPTH_MASK_64F', '_OUTPUT_ARRAY_DEPTH_MASK_8S', '_OUTPUT_ARRAY_DEPTH_MASK_8U', '_OUTPUT_ARRAY_DEPTH_MASK_ALL', '_OUTPUT_ARRAY_DEPTH_MASK_ALL_BUT_8S', '_OUTPUT_ARRAY_DEPTH_MASK_FLT', '_OutputArray_DEPTH_MASK_16S', '_OutputArray_DEPTH_MASK_16U', '_OutputArray_DEPTH_MASK_32F', '_OutputArray_DEPTH_MASK_32S', '_OutputArray_DEPTH_MASK_64F', '_OutputArray_DEPTH_MASK_8S', '_OutputArray_DEPTH_MASK_8U', '_OutputArray_DEPTH_MASK_ALL', '_OutputArray_DEPTH_MASK_ALL_BUT_8S', '_OutputArray_DEPTH_MASK_FLT', '__UMAT_USAGE_FLAGS_32BIT', '__doc__', '__file__', '__loader__', '__name__', '__package__', '__spec__', '__version__', 'absdiff', 'add', 'addWeighted', 'batchDistance', 'bitwise_and', 'bitwise_not', 'bitwise_or', 'bitwise_xor', 'borderInterpolate', 'calcCovarMatrix', 'cartToPolar', 'checkHardwareSupport', 'checkRange', 'compare', 'completeSymm', 'convertFp16', 'convertScaleAbs', 'copyMakeBorder', 'countNonZero', 'cubeRoot', 'cuda', 'dct', 'detail', 'determinant', 'dft', 'divide', 'eigen', 'eigenNonSymmetric', 'error', 'exp', 'extractChannel', 'fastAtan2', 'findNonZero', 'flip', 'gemm', 'getBuildInformation', 'getCPUTickCount', 'getHardwareFeatureName', 'getNumThreads', 'getNumberOfCPUs', 'getOptimalDFTSize', 'getThreadNum', 'getTickCount', 'getTickFrequency', 'getVersionMajor', 'getVersionMinor', 'getVersionRevision', 'getVersionString', 'haveOpenVX', 'hconcat', 'idct', 'idft', 'inRange', 'insertChannel', 'invert', 'ipp', 'kmeans', 'log', 'magnitude', 'max', 'mean', 'meanStdDev', 'merge', 'min', 'minMaxLoc', 'mixChannels', 'mulSpectrums', 'mulTransposed', 'multiply', 'norm', 'normalize', 'ocl', 'ocl_Device', 'ogl', 'patchNaNs', 'perspectiveTransform', 'phase', 'polarToCart', 'pow', 'randShuffle', 'randn', 'randu', 'redirectError', 'reduce', 'repeat', 'rotate', 'samples', 'scaleAdd', 'setIdentity', 'setNumThreads', 'setRNGSeed', 'setUseOpenVX', 'setUseOptimized', 'solve', 'solveCubic', 'solveLP', 'solvePoly', 'sort', 'sortIdx', 'split', 'sqrt', 'subtract', 'sumElems', 'trace', 'transform', 'transpose', 'useOpenVX', 'useOptimized', 'utils', 'vconcat']

哇酷小二 · 2020-11-28 08:05:05

【100元悬赏解决方案，有效期至12月9日】

听风识雨 · 2020-11-28 15:30:24

必选项:

BR2_PACKAGE_OPENCV3=y

#
# OpenCV modules
#
# BR2_PACKAGE_OPENCV3_LIB_CALIB3D is not set
# BR2_PACKAGE_OPENCV3_LIB_FEATURES2D is not set
# BR2_PACKAGE_OPENCV3_LIB_FLANN is not set
# BR2_PACKAGE_OPENCV3_LIB_HIGHGUI is not set
BR2_PACKAGE_OPENCV3_LIB_IMGCODECS=y
BR2_PACKAGE_OPENCV3_LIB_IMGPROC=y
# BR2_PACKAGE_OPENCV3_LIB_ML is not set
# BR2_PACKAGE_OPENCV3_LIB_OBJDETECT is not set
# BR2_PACKAGE_OPENCV3_LIB_PHOTO is not set
BR2_PACKAGE_OPENCV3_LIB_PYTHON=y
# BR2_PACKAGE_OPENCV3_LIB_SHAPE is not set
# BR2_PACKAGE_OPENCV3_LIB_STITCHING is not set
# BR2_PACKAGE_OPENCV3_LIB_SUPERRES is not set
# BR2_PACKAGE_OPENCV3_LIB_TS is not set
# BR2_PACKAGE_OPENCV3_LIB_VIDEOIO is not set

用qemu模拟器先跑一跑:

sudo apt install qemu-user-static -y
cp /usr/bin/qemu-arm-static output/target/usr/bin/
sudo chroot output/target/ /bin/sh

运行OK:

/ # python
Python 3.8.2 (default, Nov 15 2020, 08:10:56)
[GCC 4.8.3 20140320 (prerelease)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import cv2
>>> cv2.imread("AAAA.jpg")
>>>
>>>
>>> quit();

模拟器缺少随机数设备文件, 会提示:
Fatal Python error: _Py_HashRandomization_Init: failed to get random numbers to initialize Python

解决办法:
https://blog.csdn.net/u013538795/article/details/105703418

mknod -m 444 /dev/random c 1 8
mknod -m 444 /dev/urandom c 1 9

k455619 · 2020-11-28 15:35:43

阔以周一试一下可以的话就联系站长哦直接把钱转过来。

听风识雨 · 2020-11-28 15:36:29

如果重新勾选了某些选项没有起作用，可以用这个办法重新制作文件系统:

rm output/target -rf

#删除所有package的安装标识文件，执行make 会重新强制安装
find output/build/ |grep \\.stamp_target_installed |xargs rm -rf

make

k455619 · 2020-11-30 10:48:45

听风识雨说：

如果重新勾选了某些选项没有起作用，可以用这个办法重新制作文件系统:
rm output/target -rf

#删除所有package的安装标识文件，执行make 会重新强制安装
find output/build/ |grep \\.stamp_target_installed |xargs rm -rf

make

可以了要使用2020_02_08的buildroot 按照你上面发的设置就可以了。2020_08_02的确实不行，联系站长哦，马上转过来谢谢思密达

哇酷小二 · 2020-11-30 10:53:59

感谢楼主，感谢@听风识雨, 请加我微信 whycan_cn 领悬楼主的赏金。

听风识雨 · 2020-11-30 16:04:29

收到100元，感谢楼主，感谢小二, 感谢哇酷网。

WhyCan Forum(哇酷开发者社区)

楼主 #1 2020-11-26 10:45:03 分享评论