Abstract For robustly handling the appearance change of target object and heavy occlusion, a novel super-pixels clique based tracking algorithm is proposed. By two stage adaptive appearance modelling method, we propose the method of learning the target-background appearance framework,which is based on super pixels principle histogram bins cluster method. The process of computing superpixels cliques confidence not only store the location... For robustly handling the appearance change of target object and heavy occlusion, a novel super-pixels clique based tracking algorithm is proposed. By two stage adaptive appearance modelling method, we propose the method of learning the target-background appearance framework,which is based on super pixels principle histogram bins cluster method. The process of computing superpixels cliques confidence not only store the location information of the superpixels, the super-pixels cliques recent history and last history also are equally weighted. The first phase of two-stage adaptive cliques construct and update algorithm is target template superpixels cliques construct stage. By calculating feature distance between superpixels and cliques center, it is to determine whether a superpixel belongs to the cliques. The second phase for detection and updating stage, through compare superpixels features surrounding region of target in training frame, with cliques, the confidence of cliques can be updated. For the target appearance model adaptive learning, a principle his-togram bins clustering method be proposed to adaptive update appearance model, and the computational overhead is small. The object can be tracked under appearance changing and occlusion, by bayesian filtering method with using MAP and the adaptive appearance model. Theoretical analysis and experiments results demonstrate that our method outperforms the sate-of-the-art methods when the target under occlusion and illumination changes dramatically. Read More Read Less

Abstract World has now evolved in to a digital hub. Multimedia data such as images, videos and audio streams are being transferred, shared and accessed extensively. This extensive data is being channeled mostly through internet. Transferring multimedia data over internet securely is one of the issues faced in the digital world [1, 2]. Several traditional encryption schemes in practice are not... World has now evolved in to a digital hub. Multimedia data such as images, videos and audio streams are being transferred, shared and accessed extensively. This extensive data is being channeled mostly through internet. Transferring multimedia data over internet securely is one of the issues faced in the digital world [1, 2]. Several traditional encryption schemes in practice are not suitable now for encoding, due to higher correlation in the multimedia data. Because of this high correlation, new schemes were developed and analyzed to achieve high quality encryption for images, videos and other type of multimedia data. Several security parameters are used to assess the efficiency of the encryption algorithm. These include entropy, correlation, Peak Signal to Noise Ratio (PSNR), Unified Average Change Intensity (UACI) etc. This paper addresses the limitations faced in the encryption process presented in [3] and proposes an enhanced version to the encryption process. Read More Read Less

Abstract In this research paper, we conduct theoretical analysis and numerical analysis on novel image classification algorithm based on multi-feature extraction and modified SVM classifier. Image object classification and detection are two important basic problems in the study of computer vision, image segmentation, object tracking, behavior analysis and so on the basis of other high-level vision tasks. Existing image classification method... In this research paper, we conduct theoretical analysis and numerical analysis on novel image classification algorithm based on multi-feature extraction and modified SVM classifier. Image object classification and detection are two important basic problems in the study of computer vision, image segmentation, object tracking, behavior analysis and so on the basis of other high-level vision tasks. Existing image classification method can make full use of every single feature between the complementary characteristics of the extracted features of a large number of redundant information, which can lead to image classification accuracy is not high. For this, put forward an improved support vector machine (SVM) based on characteristics and integrated method of image classification. This method can extract comprehensive description of image content features, using principal component analysis to extract the characteristics of transformation, remove redundant information. The experimental result proves the effectiveness and feasibility of the proposed algorithm. In the final part, we conclude the paper and set up the prospect for the future research. Read More Read Less