Image classification method and system based on frequency statistical according to histogram, terminal and readable storage medium

The invention discloses an image classification method and system based on frequency according to histogram statistics, a terminal and a readable storage medium, and the method can be used in the field of dual-energy XRT (X-Ray Transformation) detection. Based on sample classification and identification of image gray frequency, effective features of the gray frequency are mined to the maximum extent, and the size of the feature space is compressed as much as possible, in particularly, an ElasticNet algorithm is adopted to obtain a feature weight matrix W, the feature contribution degree corresponding to each gray scale group is represented by using the magnitude of the weight, the features with low contribution degree are fused, and the features with high contribution degree are reserved, so that the group numbers V1 and Vh of the high and low energy gray scale groups in the gray scale histogram statistics are updated, the problems of overlarge feature space and poor classification effect in the classificatio