[Singh, 2(10): October, 2013] ISSN: 2277-9655 Impact Factor: 1.852 http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology [2780-2784] IJESRT INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY Integrated Lane Colorization Using Hough Transformation and Bilateral Filter Rajandeep Singh *1 , Prabhdeep Singh 2 *12 GIMET, Amritsar (Pb.), India rajan_gill4747@yahoo.in Abstract Lane coloration is a significant technique in a number of intelligent automobile applications, containing the lane trip recognition and warning board, intelligent journey control and autonomous driving. This paper presents a literature review on the techniques for lane coloration and explores the benefits and limits of existing lane colorization problems. It has been found that most of existing researchers has neglected the filtering and restoration techniques. However it is also found that existing researchers has also neglected the overheads of existing techniques. So in order to reduce the limitations of the existing researchers we have proposed a new strategy which uses bilateral filter as pre-processing stage which has ability to reduce the noise from images before further processing. By doing so it has started working fine even for noisy images. The proposed algorithm has been designed and implemented in MATLAB. By passing different images we have shown the significant improvement of the proposed algorithm over the existing algorithms. Keywords: Image filtering, Lane coloration, Overheads, Hough transformation, Bilateral filter. Introduction Lane coloration plays an significant role in a number of intelligent automobile applications. Various lane coloration methodologies have been proposed so far by different researchers [1] - [8]. They are classified into infrastructure-based and vision-based approaches. Although the infrastructure-based methods accomplish extremely robustness, building cost to lay leaky coaxial cables or to put magnetic indicators on the road surface is high. Vision based methods with camera on a vehicle have benefits to use well-known available lane detection in the road location and to sense a road curvature in front view. Lane coloration and lane tracking are two different steps in vision based techniques. Lane coloration is the problem of discovesring lane boundaries without any prior information of the road. Lane tracking deal with the tracking of the lane edges from frame to frame given an existing model of road geometry. Lane tracking is quite simple problem than lane coloration, as knowledge of the road geometry is known in advance which permits lane tracking processes to put properly strong constraints on the likely location and orientation of the lane edges in a new image. Lane coloration technique has to locate the lane edges without any prior knowledge of the road geometry, and do so in situations where there may be a countless clutter in the road image. This clutter can be because of the noise, dust, shadows, puddles, oil stains, tire skid marks, etc. Thus it becomes a major issue when noise is present in the input image. Thus we focus on providing an efficient algorithm which will provide better results when noise or any other unknown factor is present in the image. Literature Survey Christian et al. (2005) [1] has presented a multi-camera lane coloration algorithm that makes use of a conventional PC and a graphics card. The feature detection and lane fitting approach are able to cope with different lighting situations, weather conditions, road layouts and lane markings. Christian et al. (2005) [1] has conclude that the lane colorization is an important application in intelligent vehicular systems. Tseng et al. (2005) [2] has discussed a lane marking detection algorithm by using geometry information and modified Hough transform. First, the acquired image is divided into road part and non-road part from the geometry information. Secondly, the histogram of intensities is applied to quantize the road image into a binary image. Thirdly, a modified Hough transform method is developed to detect the lane markings in road image by using the road geometry information. However, only straight lane marking is studied here in [2]. Thus, it is interesting to develop the detection algorithm for curved lane markings.