A Novel Fish Image Retrieval Method Based on Saliency Spatial Pyramid

There are some characteristics about fish images, such as a wide variety of fish species, sheer numbers and difficult to retrieve. In this paper, we propose a fish image retrieval method based on saliency region and spatial pyramid. Firstly, we extract the interesting-regions of fish image by using saliency algorithm to reduce the influence of background in the images. And then, we extract SURF features from head, body and tail of fish respectively based on spatial pyramid to enhance the precision. At last, the similarity measurement method based on the proposed extracted features is given. In order to verify the robustness and effectiveness of the proposed method, we conduct experiments on QUT_fish_data and DLOU_fish_data datasets, the experimental results show that the method has higher recall and precision.