Open AccessObject Detection Based on Minimum Convex Hull and Generalized Hough Transform
Author(s) -
Tong Zhang,
Chenfei Song
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/768/7/072058
Subject(s) - hough transform , convex hull , feature (linguistics) , transformation (genetics) , algorithm , artificial intelligence , computation , mathematics , rotation (mathematics) , key (lock) , scale invariant feature transform , matching (statistics) , pattern recognition (psychology) , computer science , regular polygon , computer vision , image (mathematics) , geometry , chemistry , computer security , gene , linguistics , philosophy , biochemistry , statistics
In view of the low efficiency and fallibility of the feature-based matching algorithm in object detection, a new algorithm based on convex hull region and generalized hough transform is proposed.By using the region of interest to narrow the detection region and applying the key feature points to the voting elements of the generalized hough transform, the computation is greatly reduced.Firstly, the fused key feature points are extracted and filtered to generate the convex hull region which is combined with the GBR clustering algorithm to determine the region of interest. Secondly, the key feature points are taken as the voting elements in the generalized hough transform algorithm to obtain the final result. In this paper, dozens of various transformation experiments(scaling, rotation, noise, etc.) based on the proposed algorithm are compared with other similar methods, and it is verified that the method has higher efficiency and accuracy according to the theoretical basis and concrete implementation, which can meet the real-time and high-precision requirements in industrial detection.