Caption detection and enhancement choosing reasonable features according to the particularity of caption , and combining the redundant information of multiple video frame to detect and enhance caption 基于视频的字幕检测和增强根据字幕的特殊性选择合理的特征，并利用视频中多帧间的冗余信息实现字幕的检测和增强。
Following is specific work : a novel algorithm for fractal video sequence coding based on the extended quadtree video partition with the circular prediction mapping and the noncontractive mapping is proposed 提出了在扩展的四叉树视频压缩的基础上，利用环形预测映射和帧间非压缩映射相结合的视频序列分形编码方法。
To improve the origin image similarity metric , we propose a new metric called human visual system base metric ( hvs metric ) , the goal of the hvs metric is to meet the characteristic of the human visual system 针对原有帧间相似性尺度的不足，我们提出了一个符合人眼视觉特征的图像相似性尺度? hvs尺度?用于视频纹理合成。
In mrme , coding of intra - frame and inter - frame are done in wavelet decomposition domain . mrme costs much less time to perform motion estimation operation compared to block matching motion full search algorithm ( fsa ) 该方法可保证帧内、帧间编码过程始终在小波分解域内进行，运动估计运算量比块匹配全搜索法小得多。
In order to extract the contour of active image and estimate 2 - d motion effectively , a method using snakes model , where initial contour of every frame is predicted by optical flow , is given in this paper 本文提出通过计算图像序列帧间的光流变化来修正snskes动态轮廓模型，从而成功实现对二维运动目标轮廓的成功估计和跟踪。
A inter - frame error concealment algorithm based on motion estimation is proposed in this dissertation , the motion vectors and coding mode can be rebuilt , it can be applied in any frame modes and macroblock modes 本文提出了基于运动估值的帧间预测错误隐藏算法，能对运动矢量和编码模式的进行有效重建，适应各种情况的帧间错误隐藏。
In the scheme of scene control , we apply a new algorithm to accelerate the display of the scene by computing the relativity of the output frames . we use this scheme in scene display 在植物场景输出优化控制中,本文首先探讨了如何利用帧间相关性和帧内扫描线相关性来构造相关性加速算法,并采用基于图形和基于图像的方法输出场景的实验结果。
Experiments prove that the system can be used to segment and track the single moving rigid object with comparatively large inter - frame vector , and the result is good enough especially to those objects with smooth contours 实验证明，本方案对于帧间位移较为明显的单个平移运动刚体，特别是外部轮廓较为平滑的物体具有良好的分割与跟踪效果。
The center index block holds important information for the sense of human ' s eyes ; moreover , it is also the important texture information among i - frames . therefore , it will play a key role to retrieval too 由算法可知，中心索引块保留了人眼所能感知的重要信息，同时也是i帧间相关的重要纹理信息，因此，它对检索也将起到非常关键的作用。
