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.

Facial expression, a significant way of nonverbal communication, effectively conveys humans' mental state, emotions and intentions. Understanding of emotions through these expressions is an easy task for human beings. However, when it comes to Human Computer Interface (HCI), it is a developing research field that enables humans' to interact with computers through touch, voice, and gestures. Communication through expression in HCI is still a challenge. In addition, there are a variety of fields such as automotive, biometric, surveillance, teleconferencing etc. in which expression recognition system can be applied. In recent years, several different approaches have been proposed fr facial expression recognition, but most of them work only under definite environmental conditions. The proposed framework aims to recognize expressions (by analyzing the facial features extracted) based on the Active Shape Model (ASM).

Dense depth estimation is a critical application in the field of robotics and machine vision where the depth perception is essential. Unlike traditional approaches which use expensive sensors such as LiDAR (Light Detection and Ranging) devices or stereo camera setup, the proposed approach for depth estimation uses a single camera mounted on a rotating platform. This proposed setup is an effective replacement to usage of multiple cameras, which provide around view information required for some operations in the domain of autonomous vehicles and robots. Dense depth estimation of local scene is performed using the proposed setup. This is a novel, however challenging task because baseline distance between camera positions inversely affect common regions between images. The proposed work involves dense two view reconstruction and depth map merging to obtain a reliable large dense depth map.

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.

Advanced Driver Assistance System (ADAS) in combination with other active safety features like air bags etc. is gaining popularity. Vision based ADAS systems perform well under ideal lighting, illumination and environmental conditions. However, with change in illumination and other lighting related factors, the effectiveness of vision based ADAS systems tend to deteriorate. Under conditions of low light, it is therefore important to develop techniques that would offset the effects of low illumination and generate an image that appears as if it were taken under ideal lighting conditions. To accomplish this, we have developed a method, that uses local color statistics from the host image with low illumination, and enhance the same using an adaptive color transfer mechanism. By taking cues from the properties of ideal images that are saved in a database, the proposed method tends to recreate the input scene (with low illumination), into a near ideal scene, based on the database images.

In modern cars, the Advanced Driver Assistance Systems (ADAS) is cardinal point for safety and regulation. The proposed method detects visual saliency region in a given image. Multiple ADAS systems require number of sensors and multicore processors for fast processing of data in real time, which leads to the increase in cost. In order to balance the cost and safety, the system should process only required information and ignore the rest. Human visual system perceives only important content in a scene while leaving rest of portions unprocessed. The proposed method aims to model this behavior of human visual system in computer vision/image processing applications for eliminating non salient objects from an image. A region is said to be salient, if its appearance is unique. In our method, the saliency in still images is computed by local color contrast difference between the regions in Lab space.

In the research field of automotive systems, Advanced Driver Assistance Systems (ADAS) are gaining paramount importance. As the significance for such systems increase, the challenges associated with it also increases. These challenges can arise due to technology, human factors, or due to natural elements (haze, fog, rain etc.). Among these, natural challenges, especially haze, pose a major setback for technologies depending on vision sensors. It is a known fact that the presence of haze in the atmosphere degrades the driver's visibility as well as the information available with the vision based ADAS. To ensure reliability of ADAS in different climatic conditions, it is vital to get back the information of the scene degraded by haze prior to analyzing the images. In this paper, the proposed work addresses this challenge with a novel and faster image preprocessing technique that can enhances the quality of haze affected images both in terms of visibility and visual perception.

Camera sensors that are made of silicon photodiodes and used in ordinary digital cameras are sensitive to visible as well as Near-Infrared (NIR) wavelength. However, since the human vision is sensitive only in the visible region, a hot mirror/infrared blocking filter is used in cameras. Certain complimentary attributes of NIR data are, therefore, lost in this process of image acquisition. However, RGB and NIR images are captured entirely in two different spectra/wavelengths; thus they retain different information. Since NIR and RGB images compromise complimentary information, we believe that this can be exploited for extracting better features, localization of objects of interest and in multi-modal fusion. In this paper, an attempt is made to estimate the NIR image from a given optical image. Using a normal optical camera and based on the compressed sensing framework, the NIR data estimation is formulated as an image recovery problem.

Driver safety and Advanced Driver Assistance Systems (ADAS) is gaining lot of importance these days. In some countries, there are strict regulations in place which mandate the use of certain ADAS features in automobiles. However, as the need for these safety critical systems increases, the challenges associated also increase. These challenges can arise due to technology, human factors or due to nature. In countries like India, where one can expect different weather conditions with changing geography, the associated challenges are mainly due to the natural factors like haze, fog, rain and smoke. This poses a challenging problem in terms of visibility for the drivers as well as in vision based ADAS; thereby, leading to many fatal road accidents. In this paper, a novel pre-processing technique, which addresses the interesting problem of enhancing the perceptual visibility of an image that is degraded by atmospheric haze, is proposed.

Advanced Driver Assistance Systems (ADAS) have become an inevitable part of most of the modern cars. Their use is mandated by regulations in some cases; and in other cases where vehicle owners have become more safety conscious. Vision / camera based ADAS systems are widely in use today. However, it is to be noted that the performance of these systems is depends on the quality of the image/video captured by the camera. Low illumination is one of the most important factors which degrades image quality. In order to improve the system performance under low illumination, it is required to first enhance the input images/frames. In this paper, we propose an image enhancement algorithm that would automatically enhance images to a near ideal condition. This is accomplished by mapping features taken from images acquired under ideal illumination conditions on to the target low illumination images/frames.

Traffic information is very useful in planning and designing of road transport, ensuring efficient administration of road traffic, transportation agencies as well as for the convenience of road users. Traffic can be measured in terms of speed, density and flow. In this paper, we propose two different methods to measure traffic in terms of density and flow. The set up for the proposed traffic monitoring system includes a camera placed at a height from ground looking downward on the road, such that its field of view is perpendicular to the direction of motion of the traffic. The images of the road are continuously captured by the camera and processed to determine the traffic. The first method uses Gaussian Mixture Modeling (GMM) to detect vehicles. Density is calculated in terms of area occupied by the vehicles on the road. Another method of measuring the traffic flow is proposed that is based on calculation of edge points on a horizontal line drawn in the image.