Research On Feeding Patterns And Bait Technology Of Fish Culture Based On Information Of Image And Sound

As the human predation scale of marine organisms increasing,the coastal ecosystem is seriously damaged.In order to reduce the pressure of the environment,industrialized recirculating aquaculture is gradually attracting attention.In the past breeding process of various methods,the bait feeding method of timed quantification was mainly used entirely according to experience,did not take the biological feeding needs into account,resulting in excessive contamination of the water.With the rapid development of intensive aquaculture,many advanced technologies are gradually used in aquaculture,especially in the aspect of monitoring the growth process and feeding needs of creature.This paper studies a feeding pattern and baiting strategy of fish based on fusion technology of information of image and sound,in order to satisfy the demand of fish feeding conditions,and realize accurate feeding,reduce pollution and production cost.The fish feeding process has a certain rule,and baiting can be controlled by monitoring fish feeding status.Using computer image acquisition and processing system,take multiple pictures of fish feeding process at regular intervals,characteristic parameters such as area of fish,fish density,number of fish,and their changes in feeding process can be extracted through image analysis.Then establish a fish density model using improved Kmeans algorithm,a calculation model for the number of fish based on the fish image area.The results of the study showed that the curve of concentration of fish during feeding was parabolic,the number of fish in visual monitored area can be more accurately obtained based on image processing techniques,thus providing a basis for accurate feeding.Fish will produce loud sounds during feeding,the strength of the voice changes with the degree of hunger of the fish.The sound signal in the feeding process of the aquaculture fish can be obtained by using the underwater acoustic sensor,through de-noising and time-frequency analysis to obtain the time-frequency characteristics of fish feeding thus this can be used to monitor the fish feeding process and degree of hunger.Studies have shown that the sound signals of fish have a obvious change pattern,the frequency of the sound signal during the feeding process is higher than that before and after the feeding,and the power of the sound signal during the feeding process is significantly higher than the signal power of the normal time.Therefore,this can provide a basis for accurate feeding in another aspect.After information fusion based on fish image feature parameters of computer vision and image processing technology and fish acoustic feature parameters of time-frequency analysis technology,we can improve the accuracy and reliability of fish feeding monitoring.The paper adopts fuzzy theory and method to process decision-level information fusion.Taking two kinds of characteristic parameters of fish as the input,through fuzzy processing and fuzzy logic reasoning,the feeding strategy with feeding amount and feeding speed as output was obtained,which laid the foundation for the realization accurate feeding.Based on the above research work,this article carried out the overall design of recirculating aquaculture system.Initially build an intelligent breeding feeding system,and designed the appropriate control system,exploring a feasible way for the realization of green farming.