一种基于多视角视觉的漏斗形池塘精准投喂闭环控制方法

By verifying the data consistency between the multi-view imaging unit and the water quality monitoring unit within a unified control cycle, a fish school spatial occupancy map and reliable markers are generated, solving the problem of unstable feeding decisions in pond aquaculture, realizing adaptive closed-loop control, and avoiding feeding risks caused by misjudgment and abnormal data.

CN121879483BActive Publication Date: 2026-07-17ZHENGZHOU AGRICULTURAL TECHNOLOGY PROMOTION CENTER (ZHENGZHOU BRANCH OF HENAN AGRICULTURAL RADIO & TELEVISION SCHOOL)
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ZHENGZHOU AGRICULTURAL TECHNOLOGY PROMOTION CENTER (ZHENGZHOU BRANCH OF HENAN AGRICULTURAL RADIO & TELEVISION SCHOOL)
Filing Date
2025-12-31
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

In complex environments such as funnel-shaped ponds, existing automatic feeding systems in pond aquaculture struggle to establish a stable and verifiable correspondence between feeding areas, fish distribution, and water quality constraints. This leads to unstable perception, frequent misjudgments, and a lack of data alignment and consistency verification mechanisms with a unified time benchmark, resulting in unstable zoned feeding decisions.

Method used

Data is collected within a unified control cycle using a multi-view imaging unit and a water quality monitoring unit. Consistency checks are performed to generate a fish school spatial occupancy map and reliable markers. The feeding window is determined by combining water temperature, dissolved oxygen, and ammonia nitrogen monitoring data. In case of abnormal conditions, the feeding mode is switched to a limited feeding mode, forming a closed-loop control system of pre-feeding determination, feeding execution, and post-feeding evaluation.

Benefits of technology

It enables the output of stable zone feeding decisions in a traceable and reproducible manner under abnormal conditions such as communication fluctuations, occlusion, and sudden changes in brightness, avoiding the risks of overfeeding and underfeeding driven by misjudgment and abnormal data, and forming an adaptive closed-loop control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121879483B_ABST
    Figure CN121879483B_ABST
Patent Text Reader

Abstract

本发明公开了一种基于多视角视觉的漏斗形池塘精准投喂闭环控制方法，具体涉及水产养殖智能投喂控制技术领域，用于解决在池塘养殖现场多视角感知与水质监测存在波动、缺失或不一致时，分区投喂判定口径容易漂移、投喂决策可靠性不足且难以复现追溯的问题。通过在统一控制周期下将多视角成像单元与水质监测单元的采集结果按周期标识对齐，并对各视角检测结果执行一致性校核生成鱼群空间占据图与识别可信标记，且在识别可信标记不满足条件时切换为限投或停投模式，从而达到在通信波动、遮挡、亮度突变等异常情况下仍能以可追溯、可复现的判定口径稳定输出分区投喂决策，避免单视角误判或异常数据驱动的过投、漏投以及重复投喂风险。
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Intelligent pond feeding method and device based on machine vision

    CN120240380A

  • Fodder feeding efficiency evaluation method and system

    CN120598369A