Application of imagebased acquisition techniques for additive manufacturing using canny edge detection

Edge is an indispensable characteristic of an image, defined as the contour between two regions with significant variance in terms of surface reflectance, illumination, intensity, color, and texture. Detection of edges is a basic requirement for diverse contexts for design automation. This study presents a guideline to assign appropriate threshold and sigma values for the Canny edge detector to increase the efficiency of additive manufacturing. The algorithm uses different combinations of threshold and sigma on a color palette, and the results are statistically formulated using multiple regression analysis with an accuracy of 95.93%. An image-based acquisition technique system is designed and developed for test applications to create three-dimensional objects. In addition, a graphical user interface is developed to convert a selected design of a complex image to a three-dimensional object with the generation of Cartesian coordinates of the detected edges and extrusion. The developed system reduces the cost and time of developing an existing design of an object for additive manufacturing by 20% and 70%, respectively.