LU Shengyi, LIANG Dongtai, LIANG Dan, et al. Visual sorting method of mechanical parts based on Hu-SIFT feature matching and genetic algorithm[J]. 2020, (12): 123-126.
LU Shengyi, LIANG Dongtai, LIANG Dan, et al. Visual sorting method of mechanical parts based on Hu-SIFT feature matching and genetic algorithm[J]. 2020, (12): 123-126. DOI: 10.13873/J.1000-9787(2020)12-0123-04.
For the needs of rapid sorting of mechanical parts
a visual sorting method of mechanical parts based on Hu-scale invariant feature transform( SIFT) feature matching and genetic algorithm is proposed. Firstly
the part image is pre-processed by morphological processing and filtering algorithm. Secondly
an image recognition algorithm based on Hu-SIFT feature matching is designed. Hu Invariant moment and SIFT are fused to match global and local features
to extract target parts accurately. And affine transformation is used to locate the target parts. Finally
focusing on the problem of slow convergence of traditional genetic algorithms
an improved genetic algorithm based on extremal optimization is proposed
planning the optimal sorting path of manipulator target parts.The experiment results show that the average precision and recall rates of this method are 86. 06 % and 93. 25 %.The average time of optimal path planning is 3. 03 s
which can effectively improve the sorting efficiency of the manipulator.