Search Results

This paper presents a vision based pedestrian detection system. The presented algorithm is a novel method that accurately segments the pedestrian regions in real time. The fact that the pedestrians are always vertically aligned is taken into consideration. As a result, the edge image is scanned from bottom to top and left to right. Both the color and edge data is combined in order to form the segments. The segmentation is highly dependent on the edge map. Even a single pixel dis-connectivity would lead to incorrect segments. To improve this, a novel edge linking method is performed prior to segmentation. The segmentation would consist of foreground and background segments as well. The background clutter is removed based on certain predefined conditions governed by the camera features. A novel edge based head detection method is proposed for increasing the probability of pedestrian detection. The combination of head and leg pattern will determine the presence of pedestrians.

In view of the continuous efforts by the automotive fraternity, for achieving traffic safety, detecting pedestrians from image/video has become an extensively researched topic in recent times. The task of detecting pedestrians in the urban traffic scene is complicated by the considerations involving pedestrian figure size, articulation, fast dynamics, background clutter, etc. A number of methods using different sensor technologies have been proposed in the past for the problem of pedestrian detection. To limit the scope, this paper reviews the techniques involved in day-time detection of pedestrians, with emphasis on the methods making use of a monocular visible-spectrum sensor. The paper achieves its objective by discussing the basic framework involved in detecting a pedestrian, while elaborating the requisites and the existing methodologies for implementing each stage of the basic framework.

Various Advanced Driver Assists Systems (ADAS) are being used today to increase safety of drivers. These systems viz. Forward Collision Warning (FCW), Lane Departure Warning (LDW), Pedestrian Detection (PD), are all based on inputs captured using a front mounted camera. It would be useful to combine all these applications together and process the same input for different application purpose. Additionally, multicore processors are now easily available and can be used for integrating multiple ADAS applications. This would lead to reduced cost and maintenance of ADAS systems with the same performance benefits. Since current ADAS applications are sequential and/or use single core processors there is a need to parallelize these applications so that multiple cores can be utilized optimally. In this paper, we discuss our experiments and results while attempting to integrate two such ADAS applications on a multicore embedded platform.

Improving reliability and quality of software is a major aspect in automotive industry. Software reliability and quality improves by reducing bugs or defects in the software. However, finding these defects at an early stage in the software development life cycle is important to reduce rework and cost. Manually detecting defects or bugs in large code sets is time consuming and is less accurate. Hence, using static or dynamic analysis tools has become a standard practice in automotive industry. Though many such tools are commercially available, it is observed that these tools are less used for various reasons. Some of the major reasons are users need to spend considerable amount of time to learn to use these tools to get desired output reports, customized checks are required for an application that are not provided by the tool and reports are too lengthy as well as cumbersome to analyze.

This paper presents a simple yet novel approach to remove redundant data from outdoor scenes, thus finding significant application in Advanced Driver Assistance Systems (ADAS). A captured outdoor scene has two main parts, the ground region consisting of the road area along with other lane markings and the background region consisting of various structures, trees, sky etc. To extract the ground region, first the yellow and white road markings are segmented based on the HSI (Hue Saturation Intensity) color model and these regions are filled with the surrounding road color. Further the background region is segmented based on the Lab (Color-opponent) color model, which shows significant improvement as compared to other color spaces. To extract the background region such as the sky or ground region, it is assumed that the top and bottom most portions of the image does not consist of useful information.