基于Multi—Agent的图像检索技术的研究

针对当前图像特征提取方法中，算法运行速度较慢以及远程图像的信息交换中带宽的限制，设计了一种基于Multi—agent技术的图像特征提取方法，重点讨论了该方法实现的关键技术，包括纹理特征、颜色特征和形状特征的提取算法。模拟实验表明，该方法所提取的特征内容丰富，占用网络的带宽较少，采用的算法运算速度提高明显，而三种基于内容的图像的特征能满足不同类型图像的识别和检索。

Research on Image Query Technology Based on Multi-Agent

According to the slow and restrict of bandwidth in the processing of information exchange in the method of image feature picking-up,a algorithm of image feature pick-up based on Multi-Agent designed.The discuss emphasis is the key technology such as texture feature,color feature and shape feature pick-up algorithm.The simulating test proved this algorithm have advantage of abundance feature content,little-used bandwidth,high-speed.And the three content-based feature can satisfied difference type image's rec...