Study On Detection Method Of Shrimp Body Vitality Based On Machine Vision

In the process of shrimp breeding,the shrimp is greatly affected by the changes of environment factors can affect the shrimp,so it’s necessary to monitor its vitality regularly to ensure a good living environment.Taking black shell shrimps as an example,Studying the method of extacting the skeleton of shrimp based on image preprocessing.Establishing a shrimp vitality state detection system,simulating the culture environment of broodstock to detect the performance of the system based on the ECO(Efficient Convolution Operators for Tracking)target tracking algorithm and the shrimp vitality state detection system.First,in order to achieve complete segmentation of underwater shrimp target,by adjusting the Prewitt operator threshold,the test result can keep the details of the shrimp more,then removing the small target and joint expansion and operation to realize the objective of the shrimp complete segmentation,and getting the target centroid localization and shrimp skeleton extraction,it’s concluded that it’s possible to measure shrimp length for measurement based on machine vision and provide a theoretical basis.Then putting forward a shrimp target tracking method based on combine of GMM(Gaussian Mixture Mode)filter and Kalman positioning,simulating shrimp scene needs to consider the influence of water waves and shrimp body movement,combining with Kalman positioning method can measure the motion trajectory of the shrimp.In order to improve the positioning accuracy of shrimp target,proposing further the ECO target tracking algorithm.Using FHOG operator instead of HOG(Histogram of Oriented Gradients)operator to improve the tracking efficiency,this method can achieve the detection precision of each frame area,and improve the running time at the same time,comparing the manual ECO target tracking algorithm,on the basis of the centroid localization method can realize the precise measurement of shrimp movement and the instantaneous velocity.Finally,building a shrimp body vitality state detection system,and selecting reasonably hardware components based on the specific requirements of broodstock culture environment under the condition of reasonable cost control.Designing an user interface.In order to make the results more clearly represented,parameterizing the properties of the moving target to obtain the interface of the movement displacement and instantaneous velocity of the shrimp at each moment.Figure 34;Table 3;Reference 56...