Tongue image extraction method and device and computer readable storage medium

Abstract

The invention relates to artificial intelligence, and provides a tongue image extraction method, which comprises the following steps of: performing training by using an LNMF algorithm, and decomposinga matrix V corresponding to a training image into a product of a non-negative feature matrix W and a weight matrix H; wherein the dimension of the non-negative feature matrix W is n*r, r columns arefeature base images, and the non-negative feature matrix W forms a non-negative subspace; wherein the dimension of the weight matrix H is r*m; projecting the training image and the test image to a non-negative subspace respectively to respectively obtain feature coefficients, solving the similarity of the feature coefficients corresponding to the training image and the test image by using a nearest neighbor criterion, and extracting features in the test image of which the similarity is higher than a threshold value, thereby segmenting an image region representing tongue features from each testimage by using a frame. The invention further provides a tongue image extraction device and a computer readable storage medium. According to the scheme, localization of basic feature components in the original image decomposition process is further emphasized based on the LNMF algorithm, and the operation speed is higher.

Description

technical field [0001] The present invention relates to artificial intelligence, in particular to a tongue image extraction method, device and computer-readable storage medium. Background technique [0002] The existing tongue image detection method usually adopts the method of target detection, uses the sliding window to slide on the image horizontally and vertically respectively, and extracts the spatial features of the objects in the sliding window through the CNN model, and uses the SVM classifier to extract The obtained spatial features are classified to determine whether the sliding window is a tongue image. Output the coordinates of the four corners of the sliding window, and use the coordinates of the four corners to mark the position of the tongue image. However, since the size of the tongue varies greatly in different images, the angle poses also vary. Therefore, the size of the target frame is uncertain, and it is necessary to perform sliding recognition multipl...

AI Technical Summary

Problems solved by technology

Therefore, the size of the target frame is uncertain, which requires sliding recognition multiple times through target frames of various sizes, which also leads to a certain degree of complexity in target detection.

Images