Revision history [back]

the input to imdecode() should be e.g the output of imencode, bufin your previous example, not a preallocated image

please use cv::imread() , not cvLoadImage() (from outdated C api)

the input to imdecode() should be e.g a vector<uchar>, filled like a file on disk, with headers, compressed pixels, etc. (e.g the output of imencode, bufin your previous example, example ), not a preallocated an image

please use cv::imread() , not cvLoadImage() (from outdated C api)