基于兴趣点综合特征的图像检索

Image retrieval based on combined features of interest points

针对环形区域能更好地表达像素空间分布的特点,将其引入到Gabor小波纹理特征中,提出了一种基于兴趣点环形区域颜色和纹理特征的图像检索算法。首先采用自适应平滑滤波器对图像进行滤波处理,消除噪声的影响并利用快速鲁棒特征(SURF)算子检测兴趣点;然后计算兴趣点周围局部区域内环形颜色直方图及纹理特征,将其作为图像的综合特征;最后根据图像综合特征相似度,输出相似图像。实验结果表明,该算法使平均检索准确率提高至少7%。

Due to the spatial distribution expression power of the ring areas, this paper proposed a new image retrieval algorithm based on color feature and texture feature of interest points, introducing the ring area into the Gabor wavelet texture features. At first, this algorithm used adaptive smoothing filter to eliminate noises and detected interest points by the Speeded Up Robust Feature （SURF） operator. Then, it calculated the color histogram and texture feature of interest points in the ring area, as image＇s combined features. Finally, it output similar images according to the similarity of the combined features. The experimental results show that this algorithm improves the average retrieval accuracy by more than 7%.