Thorsten Graf

Survey of Pedestrian Detection for Advanced Driver Assistance Systems

Advanced driver assistance systems (ADASs), and particularly pedestrian protection systems (PPSs), have become an active research area aimed at improving traffic safety. The major challenge of PPSs is the development of reliable on-board pedestrian detection systems. Due to the varying appearance of pedestrians (e.g., different clothes, changing size, aspect ratio, and dynamic shape) and the unstructured environment, it is very difficult to cope with the demanded robustness of this kind of system. Two problems arising in this research area are the lack of public benchmarks and the difficulty in reproducing many of the proposed methods, which makes it difficult to compare the approaches. As a result, surveying the literature by enumerating the proposals one--after-another is not the most useful way to provide a comparative point of view. Accordingly, we present a more convenient strategy to survey the different approaches. We divide the problem of detecting pedestrians from images into different processing steps, each with attached responsibilities. Then, the different proposed methods are analyzed and classified with respect to each processing stage, favoring a comparative viewpoint. Finally, discussion of the important topics is presented, putting special emphasis on the future needs and challenges.

Pedestrian detection for driver assistance using multiresolution infrared vision

This paper describes a system for pedestrian detection in infrared images, which has been implemented on an experimental vehicle equipped with an infrared camera. The proposed system has been tested in many situations and has proven to be efficient and with a very low false-positive rate. It is based on a multiresolution localization of warm symmetrical objects with specific size and aspect ratio; anyway, because road infrastructures and other road participants may also have such characteristics, a set of matched filters is included in order to reduce false detections. A final validation process, based on human shapes morphological characteristics, is used to build the list of pedestrian appearing in the scene. Neither temporal correlation nor motion cues are used in this first part of the project: the processing is based on the analysis of single frames only.

3D lane detection system based on stereovision

This work presents a 3D lane detection method based on stereovision. The stereovision algorithm allows the elimination of the common assumptions: flat road, constant pitch angle or absence of roll angle. Moreover, the availability of 3D information allows the separation between the road and the obstacle features. The lane is modeled as a 3D surface, defined by the vertical and horizontal clothoid curves, the lane width and the roll angle. The lane detection is integrated into a tracking process. The current lane parameters are predicted using the past parameters and the vehicle dynamics, and this prediction provides search regions for the current detection. The detection starts with estimation of the vertical profile, using the stereo-provided 3D information, and afterwards the horizontal profile is detected using a model-matching technique in the image space, using the knowledge of the already detected vertical profile. The roll angle is detected last, by estimating the difference of the average heights of the left and right lane borders. The detection results are used to update the lane state through Kalman filtering.

A Sensor for Urban Driving Assistance Systems Based on Dense Stereovision

The urban driving environment is a complex and demanding one, requiring increasingly complex sensors for the driving assistance systems. These sensors must be able to analyze the complex scene and extract all the relevant information, while keeping the response time as low as possible. The sensor presented in this paper answers to the requirements of the urban scenario through a multitude of detection modules, built on top of a hybrid (hardware plus software) dense stereo reconstruction engine. The sensor is able to detect and track clothoid and non-clothoid lanes, cars, pedestrians (classified as such), and drivable areas in the absence of lane markings. The hybrid stereovision engine and the proposed detection algorithms allow accurate sensing of the demanding urban scenario at a high frame rate.

3D vehicle sensor based on monocular vision

Determining the position of other vehicles on the road is a key information to help driver assistance systems to increase drivers safety. Accordingly, the work presented in this paper addresses the problem of detecting the vehicles in front of our own one and estimating their 3D position by using a single monochrome camera. Rather than using predefined high level image features as symmetry, shadow search, etc., our proposal for the vehicle detection is based on a learning process that determines, from a training set, which are the best features to distinguish vehicles from non-vehicles. To compute 3D information with a single camera a key point consists of knowing the position where the horizon projects onto the image. However, this position can change in every frame and is difficult to determine. In this paper we study the coupling between the perceived horizon and the actual width of vehicles in order to reduce the uncertainty in their estimated 3D position derived from an unknown horizon.

Color image segmentation in HSI space for automotive applications

This article presents a method for classifying color points for automotive applications in the Hue Saturation Intensity (HSI) Space based on the distances between their projections onto the SI plane. Firstly the HSI Space is analyzed in detail. Secondly the projection of image points from a typical automotive scene onto the SI plane is shown. The minimal classes relevant for driver assistance applications are derived. The requirements for the classification of the points into those classes are obtained. Several weighting functions are proposed and a fast form of an euclidean metric is investigated in detail. In order to improve the sensitivity of the weighting function, dynamic coefficients are introduced. It is shown how to compute them automatically in order to get optimal results for the classification. Finally some results of applying the metric to the sample images are shown and the conclusions are drawn.

Camera Calibration Method for Far Range Stereovision Sensors Used in Vehicles

This paper presents a camera calibration method for far-range stereo-vision used for driving environment perception on highways. For a high accuracy stereovision system the most critical camera parameters are the relative extrinsic parameters which are describing the geometry of the stereo-rig. Experiments proved that even a few seconds drift of the relative camera angles can lead to disastrous consequences in the whole stereo-vision process: incorrect epipolar lines and consequently lack of reconstructed 3D points. Therefore we propose an off-line method able to give a very accurate estimation of these parameters and on-line methods for monitoring their stability in time taking into account the real-world automotive conditions, which can introduce drifts to the initial parameters due vibrations, temperature variations etc

Extracting road features from color images using a cognitive approach

This paper introduces a cognitive method for extracting significant road information (like road extents, lane markings) from mono-color images. The system is able to identify all traffic lanes and to distinguish between continuous and broken lane markings. Its output is useful in driver assistance systems (for example lane-departure warning). The cognitive aspects of the system are highlighted and the implemented algorithms are described. Finally, some results of the performed tests are introduced before drawing the conclusion.

Detection of lane markings based on ridgeness and RANSAC

Detection of lane markings based on a camera sensor can be a low cost solution to lane departure warning and lateral control. However, reliable detection is difficult due to cast shadows, vehicles occluding the marks, wear, vehicle motion, etc. The contribution of this paper is twofold. Firstly, we propose to explore another low-level image descriptor, namely, the ridgeness, instead of the gradient magnitude with the aim of getting a more reliable lane marking detection under adverse circumstances. Besides, the proposed measure comes with an associated orientation which is less noisy than the gradient one. Secondly, we have adapted RANSAC, a generic robust estimation method, to fit a parametric model to the image lane lines using both ridgeness and orientation as input data. In short, in this paper a better feature type and a robust fitting method are proposed, which contribute to improve the lane lines detection reliability, and still achieving real-time.

IR pedestrian detection for advanced driver assistance systems

This paper describes a system for pedestrian detection in infrared images implemented and tested on an experimental vehicle. A specific stabilization procedure is applied after image acquisition and before processing to cope with vehicle movements affecting the camera calibration. The localization of pedestrians is based on the search for warm symmetrical objects with specific size and aspect ratio. A set of filters is used to reduce false detections. The final validation process relies on the human shape’s morphological characteristics.