| Summary: | 碩士 === 國立成功大學 === 電機工程學系碩博士班 === 92 === As Wireless Network is now widespread,and we will use mobile device such as PDA,notebook,pocket PC to accept multimedia data ,but noise in wireless network always happened,this will damage definition of transferred image when error occurs. According to this problem we should make error resilience to raise quality of transferred image after we get it in receiver side.
In this paper,we suggest a structure of error resilience in encoder and error concealment in decoder. And we take a schema of unequal error protection,UEP ,we make groups of significant bit plane and insignificant bit plane.For significant bit plane we enhance protection to have ability of error resilience and for insignificant bit plane we use quadtree structure to compress the data.So this method has two properties ,one is error resilience and the other is small transferred data bits.Based on this method,video data is also the same as static image,and we make four partition in each image of video data,and then interleave these images to decrease error block size in one image.By this way,we can raise quality of decoding image.And then after decoder accept the image,we can first identify error condition to choose a optimal algorithm to do error concealment to get optimal quality of image.
After implementing this algorithm ,we make sure this algorithm is really a solution in wireless network ,it is adaptive to be used to transfer multimedia data in wireless network . So we can scale transferred data bits to fit the network and the mobile device .At last , because we use Haar wavelet transform , it decrease complexity of hardware ,by this we found this algorithm is really adaptive for wireless network .
|
|---|
