Font Size: a A A

Research And Design Of Network Intelligent Surveillance System Based On DM6437

Posted on:2014-02-28Degree:MasterType:Thesis
Country:ChinaCandidate:B DengFull Text:PDF
GTID:2248330398977748Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
With the rapid enhancement in substance development level of the society, people are increasingly concerned about the security of their property and themselves, along with which video surveillance system has come into people’s lives. With the evolution of video encoding technology, communication technology and semiconductor technology, the video surveillance system is moving towards the direction of digitization, networking and intelligentization. Conforming to this trend, TI introduced some new DSP products which designed for the application of networked digital media. The well performed, feature-rich products have been widely used in video surveillance industry. This paper is on the basis of this technology, and designs an intelligent video surveillance storage system which is triggered by the moving target. When moving target is detected, the system will encode the video frame automaticlly, and then will pass the data to the PC machine through network.The system consists of three parts:the analog camera, the DM6437experimental platform, and the PC. The DM6437experimental platform is the core part of the system, the hardware in the charge of the chip DM6437are for the implementation of system control and data computing. The system contains a video preprocessing chip TVP5146as the decoder which is responsible for the input analog signal converting to a digital signal, uses DDR2as an internal memory for running the applications, takes NAND Flash as external memory and adopts a KSZ8001L chip to manage the network transmission.The software of the network intelligent monitoring system is designed in the integrated development environment of CCS3.3. This design on the basis of the real-time operating system of DSP/BIOS, uses the multi-threading mechanism, dividingly completes tasks including video acquisition, video image processing, video coding and coded data transmission. The task threads synchronously communicate through semaphore. Video capture module is implemented through the GIO/FVID drive model based on DSP/BIOS. Video processing module puts forward a motion detection algorithm based the embedded system, which adopt the ViBe background subtraction algorithm to classify the pixels of all frames image in order to get the foreground object. On the design of background updating algorithm, the update policy of frame-level was introduced to make up for the defect of ViBe algorithm update the abrupt change background too slowly, which implement the real-time monitoring for moving object. When the movement happens in the video filed, the system automatically start the video encode module which encodes the video data in the H.264form. Then, the system launches the network transmission module through the Net Develop Kit. The test system show that the Network Intelligent monitoring system based on DM6437can implement moving target detection in the video filed and network transmission of video data. The whole system runs stably, and the instantaneity is very strong.
Keywords/Search Tags:Video surveillance, DM6437, Motion detection, Video encoding, NDK
PDF Full Text Request
Related items