Multiple Description Coding For Motion JPEG Video Transmission in UMTS

Abstract

In this paper, we propose a multiple description video coding system for robust transmission of video signal. We split the video frames into odd and even frames, and encode each frame separately based on JPEG standard. We assume that each description is transmitted in through WCDMA physical layer that used in UMTS. The channel is simulated based on Gilbert/Elliot’s model. We compare the result in threes cases: Video is reconstructed using only one description; video is reconstructed using both descriptions and finally video is reconstructed based on both description, and we do error concealment as well. The video quality is reported based on average peak signal to noise ratio for all cases for video sequences.多描述編碼的Motion JPEGè¦–é »å‚³è¼¸çš„UMTSæ‘˜è¦åœ¨æœ¬æ–‡ä¸­ï¼Œæˆ‘å€‘æå‡ºäº†è¦–é »ä¿¡è™Ÿçš„å‚³è¼¸å¼·å‹çš„å¤šæè¿°è¦–é »ç·¨ç¢¼ç³»çµ±ã€‚æˆ‘å€‘åˆ†å‰²è¦–é »å¹€åˆ†æˆå¥‡æ•¸å’Œå¶æ•¸å¹€ï¼Œä¸¦ç·¨ç¢¼åˆ†åˆ¥åŸºæ–¼JPEG標準每幀。我們假設每個描述是通過在UMTS WCDMAä½¿ç”¨çš„ç‰©ç†å±¤å‚³è¼¸è‹±å¯¸è©²é »é“æ˜¯åŸºæ–¼å‰çˆ¾ä¼¯ç‰¹/åŸƒåˆ©å¥§ç‰¹çš„æ¨¡åž‹ä¾†æ¨¡æ“¬ã€‚æˆ‘å€‘æ¯”è¼ƒä¸‰äº”æˆç¾¤çš„æƒ æ³ä¸‹ï¼Œçµæžœï¼šè¦–é »åƒ ä½¿ç”¨ä¸€å€‹æè¿°é‡å»º;è¦–é »ä½¿ç”¨é‡å»ºéƒ½èªªæ˜Žï¼Œæœ€å¾Œé¡¯å¡åŸºæ–¼å ©å€‹æè¿°é‡å»ºï¼Œè€Œæˆ‘å€‘åšçš„éŒ¯èª¤éš±è—ç‚ºå¥½ã€‚è¦–é »è³ªé‡æ˜¯åŸºæ–¼å¹³å‡å³°å€¼ä¿¡è™Ÿç‚ºæ‰€æœ‰çš„æƒ æ³ä¸‹å°è¦–é »åºåˆ—ä¿¡å™ªæ¯”å ±å‘Šã€‚

Keywords

• reliability. Technical feasibility