Shape-Based Matching: Application of Edge Detection using Harris Point

• Main Authors: Harun, Mohamad Haniff, Sulaiman , Marizan, Mohd Shah, Hairol Nizam
• Format: Conference or Workshop Item
• Language: English
• Published: 2011
• Subjects: TK Electrical engineering. Electronics Nuclear engineering; TA Engineering (General). Civil engineering (General)
• Online Access: http://eprints.utem.edu.my/id/eprint/11749/1/ICORAS_2011.pdf; http://eprints.utem.edu.my/id/eprint/11749/; http://tatiuc.edu.my/icoras2011/

This paper presents a sequence of object recognition algorithm using shape-based matching that mainly focused on image recognition, image segmentation, and flexible Region of Interest. The image of pyramid is used as a medium to locate each corner of the object and specified the location in details. First, the image reference is used as a training image and the template is created. Then, the process image will be compared with the reference image by using template matching to calculate the score of the successful matching. Each correspond image will be rotated around 60 degree to see whether the system able to recognize the object. All score for matching are recorded. After that, Harris point generates the specific corner of the pyramid and the location of each point is located and clarified with number starting from one. Distance between one point to another is calculated using mathematics’ equation to generate a new point between those points. The location of all generate points are displayed using Graphical User Interface (GUI). This method is proposed to develop an additional new system of the glue process in automation industry that provides input data from vision sensor to reduce possibilities of failure.