Detection method suitable for small target in road image

By optimizing the detection model through a feature complementary mapping module and a multi-kernel perceptual convolution module, the problem of insufficient accuracy in small target detection in road scenes by existing algorithms is solved, achieving higher detection accuracy and robustness, and making it suitable for road safety monitoring and traffic management.

CN121074633APending Publication Date: 2025-12-05GUANGXI YANQI INFORMATION TECHNOLOGY CO LTD +1
3 Cites 1 Cited by

Patent Information

Application Number
CN202511098743.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-08-06
Publication Date
2025-12-05

AI Technical Summary

Technical Problem

Existing object detection algorithms are not accurate enough in detecting small objects in road scenes. This is mainly because small objects have a low pixel ratio and insufficient feature resolution in the image. Traditional feature fusion methods are difficult to effectively capture cross-layer spatial semantic associations, resulting in a significant drop in detection accuracy.

Method used

The detection model is constructed using specific backbone networks and optimized modules, including a Feature Complementary Mapping (FCM) module and a Multi-Kernel Perceptual Convolutional (MKP) module. It combines NWD to replace the traditional IoU and enhances the performance of small object detection through improved loss functions and training strategies.

Benefits of technology

It significantly improves the accuracy and robustness of small target detection in road images, and can effectively handle the problems of high noise and low detail in complex scenes, providing a more efficient and accurate solution for road safety monitoring and traffic management.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the technical field of computer vision, in particular to a detection method suitable for a small target in a road image, which comprises the following steps of: collecting road image data, and marking and normalizing a road image; constructing a detection model based on a specific backbone network and module optimization; a camera image is input into a detection model for target identification and positioning, and detection information is transmitted to a monitoring center; according to the invention, the feature complementary mapping module, the multi-core perception convolution module and the optimized loss function are introduced, so that the recognition capability of the model on weak texture and small-scale targets is enhanced, the small target detection stability and accuracy under various road scene conditions can be effectively improved, and the method is suitable for large-scale detection. Therefore, the problem of insufficient detection precision of the existing target detection algorithm is solved.
Need to check novelty before this filing date? Find Prior Art