Study And Implementation Of Muti-angle And Muti-scale Whale Tail Image Retrieval System

As a part of the top of the marine food chain,cetaceans play an important role in the entire ecosystem.However,cetaceans are in an endangered situation in recent years.Therefore,the researchers need to track the cetaceans for its age,growth rate,reproductive cycle,population status,distribution,ecological environment,population characteristics and find out the causes of endangered and so on.And whale individual automatically recognition is of great significance to the scientific research work.This study searches for similar whale images based on multi-angle and multi-scale image space.Firstly,the multi-angle image space of the whale tail image is constructed,and then the local feature extraction algorithm with scale invariance is applied to the multi-angle image space,and the rotation,translation and affine invariant feature descriptors of whale tail image(ASIFT algorithm)are extracted.The main contents of this paper are as follows:(1)Observe and analyze the whale tail image dataset.Because the image in this research dataset is collected by multi-angle,the multi-angle space of input image is constructed to simulate the distortion of the target image at different angles.(2)Study on feature extraction and match method of multi-angle and multi-scale whale tail images.Based on the multi-angle and multi-scale ideas and according to the Affine-SIFT algorithm,the Affine-SURF and Affine-BSIFT feature extraction algorithms are studied in this paper.The local features of the final extracted whale tail image have affine,rotation and translation invariance,which are suitable for multi-view collection of whale taile images.(3)Selecte multi-angle multi-scale whale tail image feature extraction algorithm.The performance of Affine-SIFT,Affine-SURF and Affine-BSIFT features applied to the whale image is analyzed experimentally.After the comparison,the Affine-SIFT algorithm is used to extract the feature of the whale tail image based on the requirement of stability.The optimized Affine-SIFT is applied to the whale-tail image dataset.When the size of the similar whale image dataset is 10,the success rate SR is 0.48.(4)Design and implementation of multi-angle multi-scale whale-tail image retrieval system.Based on multi-angle and multi-scale whale tail image feature extraction algorithm and matching method,a multi-angle and multi-scale whale tail image retrieval system is designed and implemented after the requirements analysis.