Container Visual Recognition and Positioning Method Based on Color and Texture Feature Matching
Automated loading and unloading of containers is a key task in port automation,and container recognition is one of the key technologies for achieving automated loading and unloading.To address the problem of poor performance of traditional template matching algorithms under multi-scale and lighting variation conditions,an algorithm based on color and texture feature matching is proposed,which combines a color-based feature block matching algorithm with a mathe-matical morphological texture feature extraction algorithm.The algorithm uses block matching to perform preliminary position-ing of the container and uses a method to extract convex texture features on the container surface for precise positioning.The result shows that the algorithm can effectively recognize and locate containers while meeting real-time requirements and has good robustness to scale and lighting changes.
port automationcontainer recognitionblock matchingmathematical morphology
万方数据
维普
