# FishKP-YOLOv11: An Automatic Estimation Model for Fish Size and Mass in Complex Underwater Environments

## TL;DR

This paper introduces FishKP-YOLOv11, a non-contact model that accurately estimates fish size and mass in challenging underwater conditions, improving aquaculture efficiency.

## Contribution

The novel framework integrates an improved FishKP-YOLOv11 module, stereo vision, and a Random Forest model for accurate fish size and mass estimation in complex underwater environments.

## Key findings

- FishKP-YOLOv11 achieves a mean average precision (mAP) higher than various YOLO versions.
- The framework estimates fish length and width with mean absolute errors of 0.35 cm and 0.10 cm, respectively.
- Mass estimation has a mean absolute error of 2.7 g, demonstrating strong practical applicability in real aquaculture settings.

## Abstract

Fish size and mass are crucial parameters for scientific feeding and harvesting in aquaculture. This paper proposes a non-contact framework for estimating fish size and mass. The framework addresses inaccurate size and mass estimation in actual aquaculture scenarios, which are often caused by complex water quality, low illumination, and high stocking densities. The proposed framework achieves accurate size and mass estimation in complex underwater environments. The experimental results demonstrate that the framework achieves high estimation accuracy and strong practical applicability. The framework can be deployed in an actual aquaculture scenario to reduce labor requirements, decrease fish stress, and improve yield and income for producers.

The size and mass of fish are crucial parameters in aquaculture management. However, existing research primarily focuses on conducting fish size and mass estimation under ideal conditions, which limits its application in actual aquaculture scenarios with complex water quality and fluctuating lighting. A non-contact size and mass measurement framework is proposed for complex underwater environments, which integrates the improved FishKP-YOLOv11 module based on YOLOv11, stereo vision technology, and a Random Forest model. This framework fuses the detected 2D key points with binocular stereo technology to reconstruct the 3D key point coordinates. Fish size is computed based on these 3D key points, and a Random Forest model establishes a mapping relationship between size and mass. For validating the performance of the framework, a self-constructed grass carp dataset for key point detection is established. The experimental results indicate that the mean average precision (mAP) of FishKP-YOLOv11 surpasses that of diverse versions of YOLOv5–YOLOv12. The mean absolute errors (MAEs) for length and width estimations are 0.35 cm and 0.10 cm, respectively. The MAE for mass estimations is 2.7 g. Therefore, the proposed framework is well suited for application in actual breeding environments.