A Novel Classification Method of Halftone Image via Statistics Matrices

Abstract

Existing classification methods tend not to work well on various error diffusion patterns. Thus a novel classification method for halftone image via statistics matrices is proposed. The statistics matrix descriptor of halftone image is constructed according to the characteristic of error diffusion filters. On this basis, an extraction algorithm is developed based on halftone image patches. The feature modeling is formulated as an optimization problem and then a gradient descent method is used to seek optimum class feature matrices by minimizing the total square error. A maximum likelihood method is proposed according to priori knowledge of training samples. In experiments, the performance evaluation method is provided and comparisons of performance between our method and seven similar methods are made. Then, the influence of parameters, performance under various attacks, computational time complexity and the limitations are discussed. From our experimental study, it is observed that our method has lower classification error rate when compared with other similar methods. In addition, it is robust against usual attacks.

Classification of Error-Diffused Halftone Images Based on Spectral Regression Kernel Discriminant Analysis

“…where ≥ 0 is a regularization parameter and is the identity matrix. Combined with the projective function described in (14), we can easily verify that the solution * = ( + ) −1 given by (15) is the optimal solution of the following regularized regression problem:…”

“…We first extract the feature matrices of pixel pairs from the error-diffused halftone image patches, according to statistical characteristics of these patches. The class feature matrices are then subsequently obtained, using a gradient descent method, based on the feature matrices of pixel pairs [14]. After applying the spectral regression kernel discriminant analysis to realize the dimension reduction in the class feature matrices, we finally classify the error-diffused halftone images using the idea similar to the nearest centroids classifier [15,16].…”