Image processing-based recognition and extraction method for solid colored fabric defect area

The invention discloses an image processing-based recognition and extraction method for a solid colored fabric defect area, and belongs to the field of image processing. The method comprises the following steps that 101, image acquisition work on the defect area is completed; 102, noise reduction is conducted on an original image by means of median filtering, and sharpening processing is conducted through a Qualcomm Butterworth filter; 103, quick detection and recognition on the solid colored fabric defect area are conducted; 104, the image is subjected to Fourier transform; and 105, five textural feature values including the energy, the range, the variance, the entropy and the inverse difference moment are extracted and subjected to normalization processing. According to the method, the integrity of the defect area is guaranteed maximally, not only is existence of defects recognized, but also basic information including the location, the shape and the size of the defect area can be determined, it is guarante