Part recognition and location is one of the key technologies of visual picking system. The traditional identification and location algorithms have long development cycle and weak generalization ability, so it is difficult to meet the current visual picking requirements. Therefore, a part recognition and location algorithm based on convolutional neural network is proposed in this paper. This algorithm studies the hierarchical priori framework design method based on the image sensory features. On the premise of ensuring the validity of the prior boxes, the method greatly reduces the number of prior boxes and improves the speed of network reasoning. At the same time, a pyramid model of multi-mode feature fusion is studied to improve the ability of small target detection. The average detection accuracy of the network was 97.82% and the detection speed was 104.98 FPS (Frames Per Second) on the four kinds of bolt parts test sets produced. The algorithm can detect parts in real time with high precision, meet the requirements of parts detection and solve industrial practical problems.
A Hierarchical Prior Bounding Box Based on Feature Receptive Fields for Parts Object Detection
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Kapitel : 173 ; 1759-1769
2022-03-18
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Shaping receptive fields for affine invariance
IEEE | 2004
|Shaping Receptive Fields for Affine Invariance
British Library Conference Proceedings | 2004
|Remote Sensing Image Object Detection Algorithm Based on Receptive Field Enhancement
Springer Verlag | 2022
|