Arsham Abedini, Aref Miri and Alireza Maleki
Edge detection is the base of most image processing applications. There are various classical methods for performing edge detection such as canny operator. The main flaw of these methods is that they are not flexible. Pulse coupled neural network (PCNN) is proposed based on a neuron’s model in order to provide this flexibility in image processing application. This flexibility is due to the presence of many parameters which can be adjusted for different images in order to reach an acceptable performance. On the other hand, reaching an effective performance relies on specifying all of these parameters correctly which is very challenging. Due to this fact, simplified models of PCNN are presented. In this paper, we propose a parallel structure based on one simplified model in order to perform effective edge detection. Also we set this model’s parameters in a selfadaptive manner. In simulation results, we compare edge detection performance of our proposed algorithm to other methods. These results show that our algorithm has better performance in terms of noise canceling and effective edge detection.
English 
Spanish 
Chinese 
Russian 
German 
French 
Japanese 
Portuguese 
Hindi