Font Size: a A A

Rearch Of H.264 Based On Sub-Band DCT And Face Recognizition Based On HMM In Video Monitoring System

Posted on:2010-12-20Degree:MasterType:Thesis
Country:ChinaCandidate:X H LuFull Text:PDF
GTID:2178360275978221Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
In these days, video monitering system has depended more on digital and network, which provide a wider developing space for the whole security and protection industry. What's more, intelligent video monitoring system has become one of the the most important and advanced application in the security and protection area.Compared with the traditional monitering system, the intelligent mornitoring system can not only capture the video sequence but also analyse it. As a result, the monitering system can response to the abnormity target as well as the phenomena in scene, and also has the ability to forecast what would happen next, which can give the guard a warning to prepare well for the abnormal situation and minimize the lost. As it is known, the traditional mornitering system has the following lacks, such as watching all the time by guard day and night, all the solution can only been taken out after the accidents passively . However, the intelligent mornitoring system coquers these shortcomings, which is real-time, automatic, all time and preventive. So it can help even replace people in the mornitering and controlling tasks, which can lighten people's burden as well as reduce the cost. Consequencely, the intelligent mornitoring system has been widly used in all these areas such as security and protection, traffic control, military safety, fire alarm and people control and so on.In the intelligent mornitoring system, there are five key techniques, such as:video encoding technique, video network transmission technique, video storage and searching technique, motion detection and warning technique as well as object recognition and track technique. A lot of researches have been done in this area and a great of acheievements have been made in the academic world . Following on those rearches, we have made some rearch on the video encoding technique and the face recognition technique in this paper. The main work is as below:1. In the video enconding area, the new vedio encoding standard---H.264 technique has been studied and anslysed . Fast DCT is one of the key issues in H.264, according to the properties of DCT coefficients' energy distribution and the characteristics of "zigzag scan", one fast DCT algorithm is proposed based on divided subbands in the paper. Discrete Cosine Transform (DCT) coefficients of the prediction residue (Zero Quantized DCT coefficients, ZQDCT) are set zero predictably before implementing DCT and quantization (Q), and then the redundant computations are deduced greatly. One adaptive scheme is also presented with multiple thresholds to divide the subbands. By this scheme, only the DCT coefficients without ZQDCT will be computed. The experimental results show that the proposed algorithm outperforms other approaches in literature, and achieves the best performance in reducing computations in the case of the same picture quality and the same compression ratio by the traditional algorithms in H.264.2. In the face recognition area, a new approach based on lifting wavelet transform for face recognition is presented. The Daubechies 9/7 Lfting Wavelet Transform (D9/7 LWT) is employed to preprocess the normalized faces and three groups of wavelet features: approximation sub-image, horizontal detail sub-image and vertical detail sub-image, are obtained respectively. New way of extracting and using observation sequence from the three wavelet sub-images for recognition is introduced. The observation sequence is then modeled by using 1D Discrete Hidden Markov Model(1D-DHMM). The experimental results on the Olivetti Research Laboratory (ORL ) face database show that the proposed method has a high recognition rate with relatively low complexity , and it is also fairly robust to the variety of different illumination condition, face pose and expression.In this paper, we have studied the two techniques of the intelligent mornitering system, which can improve the system performance, promote the mornitering system's function, and enhance the system's level of the intelligence and networking. In one word,the work , in this paper,has a shot at the improvement on the intelligent mornitoring system.
Keywords/Search Tags:The intelligent mornitoring system, H.264 vedio encoding, Subband DCT, Face recognition, D9/7 LWT, Wavelet sub-image, 1D-HMM
PDF Full Text Request
Related items