In the process of traditional image retrieval, image features are aimed for image retrieval, once there were too much subtle features in the characteristic description, the relationship between images becomes worse, it is difficult to complete image retrieval according to the single threshold standard, reducing the efficiency of image retrieval. The massive image retrieval method based on subtle feature distinguishing, extracts color and texture features representing subtle features of image, on the basis to calculate massive multimedia image distance, in terms of color and texture features to conduct subtle features distinguish for image feature to acquire image segmentation module matrix, and establish the massive image search optimization model based on subtle feature distinguish and achieve efficient retrieval for massive multimedia image information. The experimental results show that the improved massive image retrieval model can greatly improve the retrieval efficiency and meet the actual needs of image management.
展开▼
