Fish school motion behavior parameter extraction and analysis method under breeding background condition

Abstract

The invention relates to a fish school motion behavior parameter extraction and analysis method under a culture background condition, and belongs to the technical field of aquaculture. A motion influence graph method is applied to recognition of individual features of a single target, and due to the fact that interactive motion features of target individuals are considered, the method is more suitable for object tracking of a group background; the particle advection method is adopted to realize feature recognition of multi-target individuals, and the particle advection method has low requirements on chromatic aberration when fish school feature points are extracted, so that the method is more suitable for the characteristics of low illumination, multiple targets and the like of culture, and the accuracy and practicability are greatly enhanced; aNNs are introduced into judgment of fish school motion characteristics, and the accuracy of group target tracking in the breeding monitoring process is effectively improved. Results obtained by the method can provide a basis for monitoring abnormal movement of fish schools in actual production, reducing emergencies in the breeding process and improving breeding benefits, and healthy development of the breeding industry is facilitated.

Description

technical field [0001] The invention relates to a method for extracting and analyzing fish movement behavior parameters under aquaculture background conditions, and belongs to the technical field of aquaculture. Background technique [0002] Fish behavior is an effective reference indicator to reflect the health status of fish in aquaculture. Monitoring abnormal fish behavior can provide early warnings of fish welfare status, and the process does not adversely affect the fish. By observing the behavior of farmed fish in intensive production, it can be found that the behavior of fish schools under different state conditions has different characteristics, such as parameters such as fish school speed, group dispersion and fish school spacing. Through video recording, real-time monitoring of behavior can be achieved without the need for staff to observe the breeding site. [0003] Video analysis based on traditional manual observation is difficult to apply to the background co...

AI Technical Summary

Problems solved by technology

[0003] Video analysis based on traditional manual observation is difficult to apply to the background conditions of aquaculture groups. First, the number of fish schools under aquaculture background conditions is large, the interaction frequency between individuals is high, and there is a large error close to manual observation; secondly, fish Fish behavior is an adaptive response made by individuals according to changes in the external environment and internal physiological conditions. Therefore, the behavior of fish is more complicated, and the traditional monitoring method is more difficult to operate.

In addition, the traditional manual observation process takes longer than video recording, and the efficiency of observation is low

In the prior art, there have been many studies on the application of computer vision technology to the monitoring of animal behavior, but there are few studies on the behavior of fish under aquaculture conditions, and there are few studies on the application of ANNs to the characteristics of fish movement behavior parameters. The quantitative research on the characteristics of fish movement behavior is almost blank, and the research on fish movement behavior is mostly concentrated on the individual level, not from the perspective of groups

Images