In order to reduce the complexity and weak generalization of classification method using two class support vector machines in images steganalysis, a new Kernel-based classification method using OC-SVMs with multiple hyperspheres was put forward.

This paper proposed an novel image steganalysis scheme which takes the statistical moments in the frequency domain of the histogram of wavelet subbands as features.

This paper introduces a steganalysis algorithm based on statistics method for LSB steganograhpy and mends it,proposes a method estimating the quantity of hiding data reliably.

This paper proposes a new steganographic scheme that,while keeping the advantage of low distortion of the OPA,avoids the above-mentioned peculiar colors so that steganalysis based on the exploration of these colors is defeated.

This paper proposes a novel steganalysis technique based on the statistics method for LSB insertion, in which the change of statistical characteristic caused by information hiding is used to judge the presence of secret message.

Real-time monitoring of gas-shielded tungsten arc welding pool with DSP-based image analysis;

Characteristics of pores on membrane surface by image analysis;

Talk about image analysis in fire detection;