Fish School Detection Model with Attention Mechanism
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fish school detection methods are prone to errors due to environmental noise and light interference, and have low accuracy and slow detection speeds, especially with small fish datasets and complex fish features.
Innovation Solution
A fish school detection method and system that uses a model with a feature extraction layer, a feature fusion layer, and a feature recognition layer, incorporating attention mechanisms to enhance detection accuracy by extracting and fusing feature maps, resulting in improved target detection results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor detection is used, then detection can be performed, but detection results are easily affected by noise, water quality, and light interference
Solution Approach 1:
The patent applies attention mechanism to selectively enhance important feature information while suppressing environmental interference and noise, converting the harmful effect of environmental factors into a benefit by using the attention mechanism to identify and focus on relevant features while filtering out distractions
2Measurement precision
If digital image processing method is used, then detection can be performed, but detection accuracy is low due to traditional visual algorithms and manual experience
Solution Approach 1:
The patent replaces traditional manual image processing algorithms with a deep learning-based detection model that automatically learns and extracts features from images, substituting mechanical manual processing with intelligent automated systems that achieve higher accuracy
3Measurement precision
If existing deep learning object detection methods are used, then detection can be performed, but recognition accuracy is low and detection speed is slow due to small fish datasets and complex fish features
Solution Approach 1:
The patent changes the parameters and architecture of the detection model by introducing attention mechanism and multi-branch feature fusion, which improves the model's ability to handle complex fish features and small datasets, thereby simultaneously improving both accuracy and detection speed
Data Source
AI summary
A fish school detection method and a system thereof, an electronic device and a storage medium are provided, the method includes inputting a to-be-detected fish school image into a fish school detection model; the fish school detection model including a feature extraction layer, a feature fusion layer and a feature recognition layer; extracting feature information of the to-be-detected fish school image based on the feature extraction layer, and determining a fish school feature map and an attention feature map based on an attention mechanism; fusing the fish school feature map and the attention feature map based on the feature fusion layer to determine a target fusion feature map; and determining a target fish school detection result based on the feature recognition layer and the target fusion feature map. Interference from environmental factors on detection results is eliminated, so as to effectively improve accuracy of the fish detection.