电视图象中目标检索方法

本文给出一种有效的图象目标检索方法。该方法由颜色分割和定位算法组成。颜色分割中利用样本学习和非样本交互操作，提取目标的特征颜色，并以此建立颜色索引表，以提高颜色分割速度。这位算法中利用映射技术与区域相结合的操作确定目标的大小和位置。实验结果表明，本检索算法具有正确检测和定位目标的能力。

EXTRACTING AND INDEXING OBJECTS IN VIDEO IMAGE

An effective object extracting and indexing method is proposed in this paper. This method consists of color segmentation and object location. Techniques about sample learning and nonsample learning are suggested in color segmentation, which lead to color table established. An object location algorithm is divided into statistical mapping and region-growing, which offers the location and size of it is the interesting object. The experimental results show that the proposed method have ability to extract and to index objects in the video images.