International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056 Volume: 04 Issue: 04 | Apr -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1658 STUDY OF EDGE DETECTION TECHNIQUES IN AUTOMATIC LICENSE PLATE RECOGNITION (ALPR) S.Mahalakshmi 1 , Prabha.M.Karani 2 1 Assistant Professor, Dept of ISE BMSIT&M,Bangalore ,Karnataka,India 2 Student,ISE,BMSIT&M,Bangalore,Karnataka,India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - Automatic License Plate Recognition (ALPR) can be identified as a technology which has been developed mainly based on Image Processing methodologies. It is being widely used in identifying vehicles in applications such as red-light enforcement, over speeding, bus lane control, motorway road tolling, border control and access/parking control.One of the steps in ALPR is edge detection. It refers to the process of identifying and locating sharp discontinuities in an image. The discontinuities are abrupt changes in pixel intensity scene. Edge detection is an important technique in many image processing applications such as object recognition, motion analysis, pattern recognition, medical image processing etc. Edge detectors form a collection of very important local image processing method to locate sharp changes in the intensity function. Image Edge detection significantly reduces the amount of data and filters out useless information, while preserving the important structural properties in an image. In this paper we have discussed some of the edge detection methodologies like Sobel, Canny and Harris corner algorithms and compared. The results shows the accuracy depends on the conditions of the image. Key Words: Edge detection, ALPR, Sobel edge detector, Canny edge detector, Harris corner edge detection 1. INTRODUCTION Automatic License Plate Recognition (ALPR) can be identified as a technology which has been developed mainly based on Image Processing methodologies. Basically, the License Plate Recognition (LPR) process is divided into three main parts: Plate Detection, Character Segmentation, and Character Recognition. Each of these parts plays an important role in the final accuracy. Many problems such as size variations, viewing angle, low contrast plates, vehicles high speed and time consuming algorithms have prevented researchers from introducing a single class of algorithms to solve the problem. Edges are the boundaries between object and background. It is basically the image segmentation technique which divides the spatial domain and defines the image into meaningful parts. With the help of edge detection one can easily detect the features of an image which indicates the end of one region in the image and beginning of another indicating the change in gray level. There are an extremely large number of edge detection operators available, each designed to be sensitive to certain types of edges. Edge detection is difficult to implement in noisy images, since both noise and edges contains high frequency content. Monochrome camera sensors are capable of providing higher details and sensitivity compared to color camera sensors. An IR projector increases the contrast between plateǯs characters and plates backgrounds. The IR projector is mostly useful at night to brighten the plates. We also need to synchronize the camera exposure time with the IR projector pulses in order to capture images with clearer plates. To capture individual features like lines, corners, circles and other geometric structures methods are applied to the image which also completes the distorted structures. Over the time, applications of the ALPR have spread into security establishment applications such as criminal activity monitoring and smuggling identification systems other than the widely used applications of traffic control and tolling all around the world. An important feature of the Number Plate Recognition system is to keep record of the vehicleǯs number plate image for other process if further required. If there is identification or verification process involved with application, the system will need to be connected to the respective database in order to achieve required outcome. Among these main steps, the most significant part is usually the image pre-processing step which enhances/improves the input image to a level that characters can be segmented in a correct method. Therefore, the reliability and accuracy of the ALPR systems rely on the methods that used in pre- processing. Based on the importance of the pre- processing steps used in approaches to ANPR, the aim of this paper is to compare various edge detection techniques involved in the process of plate recognition. Mainly four different edge detection filters will be used in turn by applying them on same input image. Sobel, Canny and Harris corner edge detection are compared. The algorithms are compared for accuracy as well as processing times.