Method for detecting low texture defect of wafer

The invention relates to the field of image processing, in particular to a wafer low-texture defect detection method, which comprises the following steps: based on image processing, acquiring a to-be-detected wafer surface gray level image, dividing the to-be-detected wafer surface gray level image into areas with the same size, obtaining structural similarity between each area and other areas according to a gray level mean value and a gray level variance of pixel points of each area, and calculating the low-texture defect of a wafer. Obtaining the overall similarity between each region and the adjacent region according to the structural similarity and the gray accumulated value of the pixel points of each region to obtain an overall similarity matrix, performing density clustering on the overall similarity data in the matrix, and judging whether the wafer is abnormal or not according to the clustering result. According to the method, an area corresponding to abnormal overall similarity data is obtained for a