Received: March 1, 2018 158 International Journal of Intelligent Engineering and Systems, Vol.11, No.4, 2018 DOI: 10.22266/ijies2018.0831.16 Radiographic Images Fractional Edge Detection Based on Genetic Algorithm Wessam Sayed El Araby 1 * Ahmed Hassan Madian 1,3 Mahmoud Aly Ashour 1 Ibrahim Farag 2 Mohammad Nassef 2 1 Radiation Engineering Department, Egyptian Atomic Energy Authority, Cairo, Egypt 2 Department of Computer Science, Faculty of computing and Information, Cairo University, Cairo, Egypt 3 Nanoelectronics Integrated System Center, Nile University, Cairo, Egypt * Corresponding author’s Email: eng.wessamsayed@yahoo.com Abstract: Recently, fractional edge detection algorithms have gained focus of many researchers. Most of them concern on the fractional masks implementation without optimization of threshold levels of the algorithm for each image. One of the main problems of the edge detection techniques is the choice of optimal threshold for each image. In this paper, the genetic algorithm has been used to get the optimal threshold levels for each image to enhance the edge detection of the fractional masks. A fully automatic way to cluster an image using K-means principle has been applied to different fractional edge detection algorithms to extract required number of thresholds to be optimized by the genetic algorithm. A performance comparison has been done between different fractional algorithms with and without genetic algorithm. Evaluation of the noise performance upon the addition of salt and pepper noise is measured through the peak signal to noise ratio (PSNR) and bit error rate (BER) simulated by using MATLAB. Keywords: Edge detection, Fractional systems, Soft computing techniques, Biomedical, Genetic algorithm, Clustering-Kmean. 1. Introduction Medical imaging has gained focus of many researchers as it played a very important role in the study and early diagnosis of a lot of diseases over the past five decades [1]. The medical images are mostly used as radiographic techniques to help in early diagnosis, curing and studies [2]. Nowadays, digital image processing is ameliorated by using the image enhancement techniques for additional processing [3]. Image segmentation has the goal to extract the information which is the first step in image analysis [4]. It is the method of partitioning the image to extract interest parts in a simple and easy analyzed way [5, 6]. Edge detection can be deemed as one of the most common techniques in many applications in the area of image processing such as biomedical, radiographic images. It has the goal to distinguish and locate the sharp changes in brightness of an image [7, 8]. Edge detection uses the integer-order differential methods. It could enhance the edge information effectively; however, it could be sensitive to noise and easy to lose image detail information. The fractional-order derivative has been applied to the edge detection methods to solve this problem [9]. It is still a major challenge in image processing to get the optimal threshold for each image, as these traditional techniques have limitations of using the fixed value of thresholds [10]. Soft computing as compared to the traditional techniques, it can deal with the mystery and uncertainty in image processing in a better way. It can build a machine which can work like a human to develop intelligence [11]. In this paper, the main objective is to focus on getting the optimal threshold for better edge detection through optimization technique. So, the optimization on different edge detection techniques