Sampling network has an important role in the digitization of two-dimensional image signals. Also various schemes have been proposed to take samples of two-dimensional signals, due to the characteristics of the square structure (square pixels), the use of this structure has been common in the production of digital images. In some applications, one of the structures that can be replaced with square structure is hexagonal structure. In many previous works, hexagonal structure features and its benefits have been reviewed and shown that it can be an image structure alternative to traditional square image structure. In addition, one of the most widely operations used in image processing is edge detection that is the main base of many other processing such as object detection. So in this research, at first, hexagonal structure features have been examined and then the conversion of digital images with square pixels to hexagonal pixel approaches have been reviewed and implemented. After that, three edge detection operators include canny, sobel and morphology have been implemented on the hexagonal images. Performance of edge detection operators in square and hexagonal structures are compared with each other in simulation results. Results showed that in many cases, edge detection operators in hexagonal structure have better performance than square structure and in some specific applications it can be used instead of square structures. Keywords: Hexagonal Structure Images, Square Structure Images, Edge Detection
