Video Analysis System Of Shrimp Stress Resistance Experiment Based On Computer Vision

The selection of shrimp seed with excellent stress resistance is the top priority of the shrimp aquaculture industry,and it is also the core technology for shrimp aquaculture.Researchers designed the stress resistance experiment to study the resistance limit of shrimp and find a suitable aquaculture environment.However,it cost too much time and laborious for artificial observation because of the long duration of the resistance experiment.And the movement data of shrimp cannot be calculated by manual measurement,so it is difficult to quantitatively analyze the movement of shrimp.In recent years,the rapid development of computer vision technology has been relatively mature in the detection and tracking of pedestrians and vehicles.However,there has been no application of shrimp breeding research.This video analysis system is aimed at the above problems and based on computer vision technology and OpenCV and Qt designed a video analysis system for the analysis of shrimp resistance experiment video.The main work is as follows:(1)Target detection and separation of shrimps in the video of stress resistance experiment.In this video analysis system,the foreground target and the background image are distinguished through the establishment of the stress video background model,based on the Background Modeling Detection technology.And the detected foreground targets are separated by filtering and thresholding processing.(2)Judging the shrimp survival status and recording the dead time of the shrimp in the video of stress resistance experiment.In this video analysis system,the HOG feature extraction of the target detection results is put into the SVM to classify the shrimp survival status.And it is determined whether the shrimps have died by detecting whether the shrimps are in a state of being rolled and bent for a long time.(3)Single-target tracking of the shrimp in the video of stress resistance experiment.The results of the detection algorithm are used to initialize the Camshift tracking algorithm and find the lost target to do the target tracking of the shrimp in the video.(4)Computation of motion data of shrimp in the video of stress resistance experiment.The position of the shrimp in each frame can be obtained according to the target tracking result.We calculated the data of shrimp movement such as the movement distance,speed,acceleration and angle of movement.(5)Using Qt to design a simple interface.Users can input and output files,set parameters and perform real-time data processing through the interface,which facilitates the operation of non-computer professional shrimp aquaculture researchers.This video analysis system applies computer vision technology to the experimental study of shrimp stress resistance,which not only saves the manpower and time of researchers involved in shrimp farming,but also provides more accurate data for quantitative analysis,providing important assistance for shrimp seedlings breeding and prawn behavior research.