Vision-Based For The Recognition And Identification Of The Edge Of A Tooth Saw Butt Joint Shape

Lagani, Muhammad Shadiq (2019) Vision-Based For The Recognition And Identification Of The Edge Of A Tooth Saw Butt Joint Shape. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) Vision-Based For The Recognition And Identification Of The Edge Of A Tooth Saw Butt Joint Shape.pdf - Submitted Version Download (1MB)

Abstract

Nowadays, the most well-known autonomous method in recognizing and identifying the edge of a butt joint implements the usage of vision sensor or laser-assisted vision sensor because of its performance and robustness compared to a manual approach. The most common vision sensors that are used are charge-coupled device (CCD) and complementary metal-oxide semiconductor (CMOS) cameras. This research paper presents the development of a vision-based method to recognize and identify the edge of a tooth saw butt joint shape and evaluation of the accuracy and repeatability of the proposed method. CMOS camera is used in this research because of its high readout speed and inexpensive cost over the CCD camera. The methodology for the digital image processing of the recognition and identification of the edge of a tooth saw butt joint shape comprises of four processes: (1) image pre-processing (2) image segmentation (3) morphological image processing (4) edge butt joint feature points representation and description. The feature points of the edge of a tooth saw butt joint shape which is the start point, supporting point 1 & 2 and the end point is in x and y coordinates of the pixel value image captured by the camera. All the variables in the image processing such as, the threshold values for the edge detection techniques to convert the original image to binary image, the size of the structuring element of the morphological operation dilation and the minimum quality for corner detection is determined and compared to find the most suitable value. The process of the recognition and identification of the edge of a tooth saw butt joint shape is done using different edge detection techniques such as Sobel, Prewitt, Roberts and Canny edge detection technique. The average readings for the feature points is compared to the original points. The comparison shows the accuracy of each method. The average readings are used to calculate standard deviation to show each method’s repeatability. The findings suggest that Canny edge detection technique is the most accurate method and all the techniques has a high repeatability due to the reliability of the variables determined and procedures done.

Actions (login required)

View Item