1Department of Computer, Ardabil Science and Research Branch, Islamic Azad University, Ardabil, Iran. Department of Computer, Ardabil Branch, Islamic Azad University, Ardabil, Iran
2Department of Mathematics, Ardabil Branch Islamic Azad University, Ardabil, Iran
3Department of Computer, Ardabil Science and Research Branch, Islamic Azad University, Ardabil, Iran Department of Computer, Ardabil Branch, Islamic Azad University, Ardabil, Iran
Image recognition is one of the most important field in image processing that in recent decades had much attention .Due to expansion of related fields with image processing and various application of this science in machine vision, military science, geography, aerospace and artificial intelligence and lots of other aspects, out stand the importance of this subject.One of the most important aspects of this case is line detection that it could be base of image recognition.Because all pictures, shapes and even whole written in picture has been formed from lines. Accordingly, exact line detection can be basis to exact image recognition.One of the methods that are used in this aspect is to identifying lines, using Hough transform .This technique is one of useful and practical method in the field of line detection. Problem with this approach is the errors in detected lines that with improving this method can lead thy way for improvement of lots of other application in this topic.In this thesis a method for improving the Hough transform in line detection is proposed. According to the achieved results it can be seen the effect of proposed method in improving Hough transform.