A YOLOv7 target recognition method and system based on global coordinated attention

By introducing global coordinated attention and multi-detection head modules into the YOLOv7 model, the problem of low efficiency and low accuracy in water bleaching garbage identification is solved, and high-precision detection of small-sized water bleaching garbage is achieved to reduce missed inspections. Improved recognition accuracy.

CN118447235BActive Publication Date: 2025-01-14GUANGZHOU ZHONGYIYONG INTELLIGENT TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410628490.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-21
Publication Date
2025-01-14
Estimated Expiration
2044-05-21

AI Technical Summary

Technical Problem

In the process of identifying water bleach garbage, the existing technology has problems such as low recognition efficiency, low accuracy, large dependence on image quality, frequent misjudgment of similar objects and weak generalization ability, especially in complex riverside environments, light and water waves. Severe interference leads to insufficient extraction of small-sized rinse garbage features, resulting in missed identification.

Method used

Using the YOLOv7 target recognition method based on global coordinated attention, by introducing global coordinated attention module and multi-detection head module, a water waste garbage image recognition detection model is built to improve feature extraction accuracy and water waste garbage detection accuracy of different sizes to avoid missed inspection and misjudgment.

Benefits of technology

It effectively improves the feature extraction accuracy of the image recognition and detection model of wafting garbage, reduces missed detection, improves the detection accuracy of wafting garbage of different sizes, and meets the identification needs in complex environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118447235B_ABST
    Figure CN118447235B_ABST
Patent Text Reader

Abstract

The present application discloses a YOLOv7 target recognition method and system based on global coordinated attention, the method comprising: obtaining a water-flooded garbage image and performing data annotation processing to obtain an annotated water-flooded garbage image; based on the YOLOv7 model, introducing a global coordinated attention module and a multi-detection head module to construct a water-flooded garbage image recognition and detection model; inputting the annotated water-flooded garbage image into the water-flooded garbage image recognition and detection model to perform image recognition and detection processing to obtain a water-flooded garbage recognition result. The embodiments of the present application can improve the feature extraction accuracy of the water-flooded garbage image recognition and detection model and improve the detection accuracy of water-flooded garbage of different sizes. The present application can be widely used in the field of image recognition technology.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Rubber sealing ring defect detection method based on YOLOv7-tiny

    CN117830253A