PDF Review: Image Processing Road Detection - IRD India

(1)

International Journal on Advanced Computer Theory and Engineering (IJACTE)

_______________________________________________________________________________________________

ISSN (Print): 2319-2526, Volume -4, Issue -1, 2015 22

Review: Image Processing Road Detection

1Pranjali Ulhe, ²Prajakta Bhure, ³Ankita Bakade, ⁴Ashwini Bute, ⁵Najuka Koram

1,2,3,4,5Department of Computer Engineering, S.D.C.O.E, Wardha , Maharashtra, India

Email: ¹[email protected], ²[email protected], ³[email protected], ⁴[email protected],

5[email protected]

Abstract -This paper presents a review of recent vision based on-road detection systems. As we know most of the Indian rural and sub urban roads are not ideal for driving due to faded lanes, irregular potholes, improper and invisible road signs. This has led to many accidents causing loss of lives and severe damage to vehicles. Many techniques have been proposed in the past to detect these problems using image processing methods. But there has been little work specifically carried out for detecting such issues of Indian roads. To address this acute problem, the study is undertaken with the objectives like, to make a survey of Indian roads, to suggest the method to detect lanes, potholes and road signs and their classification and to suggest automated driver guidance mechanism.Our focus is on systems where the camera is mounted on the vehicle and also being fixed such as in traffic/driveway monitoring systems.

In this regard, Hough Transformation method is adopted for Lane detection, where as Color Segmentation and Shape Modeling with Thin Spline Transformation (TPS) is used with nearest neighbor classifier for road sign detection and Classification. Aiming to quickly hypothesize the location of obstacle in an image as well as to verify the hypothesized locations are reviewed next.

Keywords:Image processing; obstacle-recognition; matlab;

road image analysis; hough transform, segmentation;

I. INTRODUCTION

Developing an automated driver guidance system is very important in the context of Indian road conditions. A driver finds it difficult to control the vehicle due to sudden pot holes or bumps or sudden turns where the road signs are not very prominent or missing most of the times.

Suppose if there is a system with integrated motion camera and an integrated onboard computer with the vehicle, a simple driver guidance system based on frame by frame analysis of the motion frames can be developed and there by generate the alarm signals accordingly. So that the driving can be made quite easier. Road Image analysis is very important aspect for automated driver support system. Real-time qualitative road data analysis is the cornerstone for any modern transport system. So far, most of the analysis is done manually and the use of image processing techniques for qualitative analysis is still at its early stage. In this paper description about novel image

processing algorithms together with the results is given, which assign a qualitative description to a road scene.

II. SURVEY ON ROAD SCENE

The qualitative description of a road scene can be used for controlling road lights and putting hazard signals on the road side, thereby warning drivers to slow down or direct them to alternative routes. An attempt has been made to analyses a wider view of the path and evaluate the whole description of road status as shown in below fig.1

FIG.1 Road condition Scene

The study presents an application of computer vision methods to traffic flow monitoring and road traffic analysis. The said application is utilizing image-processing and pattern recognition methods designed and modified to the needs and constraints of road traffic analysis. These methods combined together gives functional capabilities of the system to monitor the road, to initiate automated vehicle tracking, to measure the speed, and to recognize number plates of a car.

Software developed was applied, approved with video monitoring system based on standard CCTV cameras connected to wide area network computers. Traffic signal lights are triggered using an inductive loop. At a traffic light, an automobile will be stopped above an inductive coil and this will signal a green light. Unfortunately, the device does not work with most motorbikes. Using a passive system such as a camera along with image processing may prove to be more effective at detecting vehicles than the current system. The study determines that the features of various motorbikes and automobiles are sufficient enough to classify it as traffic. The study introduces a visual zebra crossing detector based on the Viola-Jones approach.

(2)

_______________________________________________________________________________________________

The study proposes a methodology to detect lanes in video frames. This method used here is a parabolic lane model to represent lanes in each video frame.

Randomized Hough transform and a Genetic Algorithm is used to estimate the parameters of lane model. The proposed method is tested on different road images taken by a video camera from Ghazvin -Rasht road in Iran. This paper presents road signs are detected by means of rules that restrict color and shape and require signs to appear only in limited regions in an image. Which are then recognized using a template matching method and tracked through a sequence of images. Real-time Traffic Sign Detection paper Yield sign, stop sign and red-bordered, circular signs are considered. First, image is color segmented based on a thresholding technique. Then, corner features are detected using convolution masks Geometric constraints used for shape recognition along verification methods for each sign. Hough has proposed an interesting and computationally efficient procedure for detecting lines in pictures. In this paper the use of angle- radius rather than s lope-intercept parameters simplifies the computation further. Also it is concentrated on general curve fitting, and gives alternative interpretations that explain the source of its efficiency.

III. IMAGE ANALYSIS:

The images used in this research were acquired using stationary traffic cameras mounted at various locations in the metropolitan Chicago area. These 2D images used JPG format and were represented using a two dimensional array such that each value in the array represented the RGB color triplet for an individual image pixel. This section details the techniques used to enhance the traffic images and extract the road direction and vanishing point for use in heading detection.

3.1 GRAYSCALE CONVERSION:

The first step in image analysis was the conversion of the RGB image into grayscale using standard techniques. An example grayscale image is shown in Figure.

FIG. 2 image obtained from traffic camera.

FIG. 3 grayscale converted image.

3.2 MEDIAN FILTERING:

After grayscale conversion, the noise within the image was smoothed using a median filter (see Figure 3). A median filter modifies an image pixel such that the new pixel value is selected as the median of the neighboring values within a range. This filtering has the effect of removing random noise in images while still maintaining the overall integrity of image regions and boundaries.

This type of filter is typically categorizes as a blurring filter in many image processing toolkits.

3.3 THRESHOLDING:

The next step in processing involves the conversion of the grayscale image into a black and white image through a process called Thresholding. This process involves the selection of a threshold value in the grayscale range which acts as a cutoff for determining which pixels should be converted to white or black in the resulting image. The resulting black and white image is shown in Figure 4.This image shows a clear perspective flow of image lines from the lower right corner toward the upper left vanishing point.

FIG.4 Image after median filtering.

FIG.5 The thresholded black and white image.

IV. METHOD DESCRIPTION:

4.1 Image processing:

Digital Image Processing converts image data into digital data in order to bring information through various processes in a computer. The computer system, input or output data needs to be in digital format only. In the digital image processing, the system receives image data, then calculates it, and finally output digital image data. To store image data into the computer's memory can be performed by reservation the memory of the machine in the form of an array. The value in each array represents the quantity of pixel. The position of the image is determined by the position of the array.

(3)

_______________________________________________________________________________________________

4.2 Hough Transform:

The Hough transform is a popular technique which can be used to isolate features of a particular shape with in an image. In this work, it is used for detecting line road as lane detection. There are 2 methods for computing the Standard Hough transform (SHT) of the binary image BW, which is an algorithm of parameter matrix whose rows and columns correspond to rho (ρ) and theta (θ)

values respectively. The first method is given below:

FIG. 6 Detecting line road by using the Hough transform.

V. PROPOSED WORK:

A sample block diagram of the proposed work is represented in figure as shown below.

FIG. 7 Generalized block diagram of the proposed work In this study a database of Indian road images is collected.

Images are acquired from a sedan from outside the driver‘s window with a digital still camera from a stationary vehicle. Hence, the images give an estimated view of the road side as seen by the driver. The proposed work considers stationary images to build the image processing system and leaves the blur removal filtering for future enhancement in the work. Acquired images are fed to the image processing system. Generalized hough transformation is applied over the image to mark the lanes. As the lanes in some of the roads are found to be not clear, a threshold based image enhancement is adopted.

Many works like Road Traffic Analysis and Traffic Sign Detection were conducted. Here, TPS is applied over the extracted sign image to get features for road signs. These features are classified with Nearest Neighbors Classifier

to classify the road signs if any, present in the scene. For pothole detection K-Means clustering based segmentation is applied over the scene. Pot holes present distinct change in the texture of the road. Hence areas with pot holes are segmented as independent unit.

VI. CONCLUSION

Real-time qualitative road data analysis is the cornerstone for any modern transport system. So far, most of the analysis is done manually and the use of image processing techniques for qualitative analysis is still at its early stage.

In this paper description about novel image processing algorithms together with the results is given, which assign a qualitative description to a road scene. The qualitative description of a road scene can be used for controlling road lights and putting hazard signals on the road side, thereby warning drivers to slow down or direct them to alternative routes. This full frame image processing application requires a low-cost frame grabber and a Pentium-based computer system for on-line real-time operations. We have presented a survey of vision-based on-road obstacle detection systems—one of the most important components of any driver assistance system.

REFERENCES:

[1] M. Xie, L. Trassoudaine, J. Alizon, J. Gallice, Road obstacle detection and tracking by an active and intelligent sensing strategy, Machine Vision and Applications 7 (1994) 165–177.

[2] Matthews N, An P, Charnley D, Harris C.:

‗Vehicle detection and recognition in grayscale imagery‘, Control Eng. Pract. 1996, 4, 473–479.

[3] M. Xie, L. Trassoudaine, J. Alizon, J. Gallice, Road obstacle detection and tracking by an active and intelligent sensing strategy, Machine Vision and Applications 7 (1994) 165–177.

[4] A. Broggi, S. Berte, Vision-based road detection in automotive systems: a real-time expectation-driven approach, Journal of Artificial Intelligence Research 3 (1995) 325–348.

[5] M. Bertozzi, A. Broggi, GOLD: a parallel real-time stereo vision system for generic obstacle and lane detection, IEEE Transaction Image Processing 7 (1) (1998).

[6] Singh, P. P., and Garg, R. D., 2012. Automatic road extraction from high resolution satellite images using adaptive global thresholding and morphological Operations. Journal of the Indian Society of Remote Sensing, 41(3), pp. 631-640.

(4)

_______________________________________________________________________________________________

[7] Zielke T, Brauckmann M, and Seelen WV:

‗Intensity and Edge- Based Symmetry Detection with an Application to Car-Following‘. CVGIP:

Image Understanding, 1993, 58,177-190.

[8] R. O. Duda and P. E. Hart, ―On the Use of the Hough Transformation to Detect Lines and Curves in Pictures,‖ Comm.ACM, vol. 15, no. 1. pp.

11-15, January 1972.

[9] Jones W.: ‗Keeping Cars from Crashing‘, IEEE Spectrum, 2001, 38(9): 40-45.

[10] Bertozzi M and Broggi M.: ‗Vision-Based Vehicle Guidance‘. Computer, 1997, 30(7),49-55.

[11] Tsugawa S.: ‗Vision-based vehicles in Japan:

machine vision systems and driving control systems‘, IEEE Transactions on Industrial Electronics, 1994, 41(4), 398–405.

[12] B. Fazenda, H. Atmoko, F. Gu, L. Guan1 and A.

Ball, ‖Acoustic Based Safety Emergency Vehicle Detection for Intelligent Transport Systems,‖

ICCAS-SICE, Fukuoka, Aug 2009, pp.4250-4255.

[13] M. Fathy and M. Y. Siyal ―A window-based image processing technique for quantitative and qualitative analysis of road traffic parameters‖

