Research of Image Retrieval Method Based on Interest Points
Wei Hai-fang · Computer Era · 2010
A new image retrieval method combining many image features based on interest points is presented.The method extracts texture features as the main characteristics to depict image by using the annular color histogram around the local area of interest points and Gabor wavelet transformation,and retrieves image by combining the space distribution of interest points.It not only overcomes the defect not able to truly characterize image by using single feature,but also ensures the recognition invariance of retrieval algorithm for image rotation and translation.Compared with the similar methods,this method can effectively improve the accuracy rate of image retrieval.