Ocean farm fishery resource conservation effect evaluation method based on bionic eels

Abstract

The invention discloses an ocean farm fishery resource conservation effect evaluation method based on bionic eels. The method comprises the following steps: S110, acquiring image information and environment information of fishes in an ocean farm via the bionic eels; S120, according to the image information of the fishes, constructing a neural network model via training samples to recognize and classify the fishes in the obtained image information according to species; S130, estimating the quantity of the fishes in a shoal of fish in the ocean farm via a target tracing algorithm according to the image information of the fishes and a recognition and classification result; and S140, evaluating a conservation effect on ocean farm fishery resources according to the recognition and classification result of the fishes, the fish information and the environment information. According to the method, the conservation effect on the ocean farm fishery resources can be efficiently and accurately evaluated; flexible activity in each sea area can be achieved, and consumption of various resources can be reduced, disturbance to the marine environment is reduced, and the marine resources is protected.

Description

technical field [0001] The present invention relates to biological classification and an evaluation method for marine pastures, and more specifically, relates to a method for evaluating the conservation effect of marine pasture fishery resources based on the classification of organisms in the image processing field. Background technique [0002] At present, marine ranching is in a stage of rapid development, and the basis for the development of marine ranching is the assessment of fishery resources in marine ranching. [0003] When assessing fishery resources in marine pastures, fishing sampling surveys are traditionally used, which will consume a lot of manpower, material resources, and financial resources, and the entire survey period is long, and only the approximate number of fish can be estimated; the patent application number is CN201911053994.2 A device and method for evaluating the biomass of deep-sea culture cages based on statistics. The classification method adopt...

AI Technical Summary

Problems solved by technology

[0003] When assessing fishery resources in marine pastures, fishing sampling surveys are traditionally used, which will consume a lot of manpower, material resources, and financial resources, and the entire survey cycle is long, and only approximate fish populations can be estimated; the patent application number is CN201911053994.2 A device and method for evaluating the biomass of deep-sea culture cages based on statistics. The classification method adopted is that the fish are classified into N categories according to the size corresponding to the quality, and the total mass of the fish in the cage is counted by statistical methods, which cannot be identified. The accuracy of the evaluation of biological species and the number of species is not enough, and the use of neural network-based classification models and target tracking to count the number of fish can intuitively reflect the growth status and resource status of aquaculture species in aquaculture waters. When using resources, the noise is loud, the environment is greatly disturbed, and it will cause harm to the fishery resources themselves.

Images