基于块边缘特征直方图的图像检索

An Image Retrieval Algorithm Based on Block-Edge-Pattern Histogram

本文提出一种利用块边缘特征来进行图像检索的方法。首先对图像中每个8×8的块分别提取其决边缘特征,利用块边缘特征构建三维或四维直方图进行图像检索。所有数据在 DCT 压缩域中进行处理,从而降低了计算复杂度。实验结果表明,该方法与传统的基于 DCT 压缩域的图像检索方法相比,具有检索准确度高和检索速度快的特点。

In this paper, we propose a new image retrieval algorithm based on block edge patterns. Block edge patterns are extracted respectively from each 8×8 block of image. Then a 3-D or 4-D histogram is constructed by utilizing block edge patterns to retrieve images inside image databases. Because the edge orientation of each block can be estimated in discrete cosine transform （DCT）, it will reduce the complexity of evaluating edge orientation. In comparison with ex isting representative techniques, the experimental results show the superiority of the proposed algorithm in terms of re trieval precision and processing speed.