小波变换域内基于方向梯度相角直方图的图像检索算法

利用反对称双正交小波所具有的多尺度边缘表征能力，基于图像边缘的梯度相角构建统计特征量用于反映图像的内容并支持检索，实验结果表明该算法不仅具有较高的检索效率，同时对图像中存在的光照变化和几何变化（尺度、平移和旋转等）具有较强的鲁棒性，且用于检索的特征量能够直接基于部分解码的小波压缩数据得到。

Wavelet Domain Image Indexing Using Directional Gradient Angular Histogram

An image-indexing technique using directional gradient angular histogram of image edges is proposed in anti-symmetrical bi-orthogonal wavelet domain. Analysis and experimental results reveal that the proposed technique not only has a superior image-retrieving performance, but also the robustness against illumination changes, position translation, scale and rotation of the image content. In addition, the proposed technique is easily integrated with a wavelet-based image coder since the angular histogram features can be constructed directly on the quantized wavelet coefficients in compressed domain.