The Research And Inplementation Of Tree Species Retrieval System Based On Hadoop

The computer-based retrieval system has been used as an essential tool by foresters to recognize tree species.Since there are a wide variety of existing trees,the retrieval time is an important indicator of the syste performance.According to the review on the existing retrieval systems,it is found that the overall system architecture,techniques for system development and the proposed algorithm can affect the system significantly.So far,the speed of the current retrieval system and the efficiency of batch retrieval are still need to be improved.The thesis proposes a solution for the above-mentioned problems which uses Hadoop distributed technology to retrieved tree species.Correspondingly,the thesis focus on the study of the Hadoop scheduler and a Hadoop scheduler based on the colony algorithm is developed.This system can achieve good performance on allocating resources and scheduling tasks in the cluster where the nodes are different from each other.It makes the system more efficient and reduces the response time.The system is able to retrieve all tree species at a time and return the results to the users.The user can query the results from the foreground system.The proposed system can solve the problems of fast retrieval and batch retrieval of tree species.