This paper deals with a problem of road tracking and vehicle positioning based on sequential images captured by a vehicle mounted monocular colour camera. The paper presents a new region-based road tracking method, wh...
详细信息
This paper deals with a problem of road tracking and vehicle positioning based on sequential images captured by a vehicle mounted monocular colour camera. The paper presents a new region-based road tracking method, which is suitable for both structured and unstructured roads. The proposed method is based on our original road segmentation algorithm and estimates road model parameters from measurements obtained by analysis of segmented road region. Although the proposed method is tested over a limited range, evaluation on real image sequences shows promising results for various types of road.
The goal of this research is to develop an algorithm that accurately segments high-resolution images, where linear features, such as roads, are corrupted by noise. In high-resolution images, there are two types of noi...
详细信息
The goal of this research is to develop an algorithm that accurately segments high-resolution images, where linear features, such as roads, are corrupted by noise. In high-resolution images, there are two types of noises that are obstacles to road segmentation. Noise could be within road areas (such as cars, water, differences in composite road surface mix) or unwanted contents outside road areas, like buildings and trees. To remove unwanted contents, the geographical information from the United States Geographical Survey (USGS) is used. The USGS provides a collection of road centre line information that has been collected for many years and can be used to limit the area for road segmentations close to roads. In this paper, a standard process was developed to align the USGS geographical information with the high-resolution images. USGS geographical data is used to eliminate background clutter that is disjunct from roads. The road segmentation process is then reduced to dealing with automobile traffic, shadows and pavement colour discontinuity within road areas. In order to achieve reliable road segmentation in the presence of these objects, the mean-shift clustering approach is used within the hue-saturation-intensity (HSI) space. Conditional morphological image processing techniques are also used to significantly improve the segmentation results. The proposed method results in the average accuracy of road segmentation above 85%.
A robust method of road segmentation for Autonomous Land Vehicle (ALV) navigation system is presented. The Main contribution of the present road segmentation method Consists of an effective improvement on the mean shi...
详细信息
ISBN:
(纸本)9781424410651
A robust method of road segmentation for Autonomous Land Vehicle (ALV) navigation system is presented. The Main contribution of the present road segmentation method Consists of an effective improvement on the mean shift algorithin dedicated to road segmentation and an extension to the Bayesian method due to its suffering from incorrectly predicted edge and the non-generalization from the sampled pixels to the unsampled pixels. The improved mean shift algorithm transforms the road image to get better feature representation which highlights the intrinsic characteristics of road images. road images are segmented by fusing with the results gotten by the improved mean shift algorithm and the extended Bayesian method, the scene variation information between adjacent frames, and the vehicle motion information. The method needn't assume a simplified road model and overcomes the shortcomings brought out by it. The experimental results show that the method has good performance and increases the segmentation accuracy to an extent.
Considering the requirements of real time and robustness for the segmentation algorithms for mobile robots, an optimal threshold segmentation algorithm based on mathematical morphology was studied. Two improved measur...
详细信息
ISBN:
(纸本)9780769538594
Considering the requirements of real time and robustness for the segmentation algorithms for mobile robots, an optimal threshold segmentation algorithm based on mathematical morphology was studied. Two improved measures, i.e., introducing an improved combination filter of median and multi-scale, and increasing morphological opening and closing operation, were proposed for the general optimal threshold segmentation algorithm. The experimental results indicated that the algorithm can not only accurately segment road and off-road regions within the unstructured road images that contain three common types of noise, but better overcome the influence of the road shadow, water marks, cracks and damage on the segmentation.
In this paper, we present a novel algorithm to extract road boundary for autonomous vehicles by analyzing road boundary in a video sequence captured from a moving vehicle. More specifically, the proposed algorithm 1) ...
详细信息
ISBN:
(纸本)9780889865839
In this paper, we present a novel algorithm to extract road boundary for autonomous vehicles by analyzing road boundary in a video sequence captured from a moving vehicle. More specifically, the proposed algorithm 1) extracts the edge map of an input image, 2) deletes unnecessary edge components that do not belong to road boundary and merges broken edges of road boundary, and 3) removes shadows that lead false positive of road boundary. In addition, extracted edges from multiple images, i.e., a block of images, can be successfully combined to detect changes of road boundary, making possible to notice that an intersection is located ahead. Experimental results are provided to show the robustness and effectiveness of the proposed algorithm under various road conditions such as shadows and snow, etc.
A road boundary detection method based on statistical test is proposed in which edge location and edge orientation are considered. The evidence of global boundaries on the road image can be obtained by means of statis...
详细信息
ISBN:
(纸本)0780372689
A road boundary detection method based on statistical test is proposed in which edge location and edge orientation are considered. The evidence of global boundaries on the road image can be obtained by means of statistical model analysis. Through statistical test, a multi-order line segment description of the road boundaries is achieved. The results are satisfied when this method is used in structured road image with dim shadows.
In this paper an integrated vision system for autonomous land vehicle is described. The vision system includes 2D and 3D vision modules and information fusion module. The task of 2D vision is to provide the physical a...
详细信息
In this paper an integrated vision system for autonomous land vehicle is described. The vision system includes 2D and 3D vision modules and information fusion module. The task of 2D vision is to provide the physical and geometry information of road, and the task of 3D vision system is to detect the obstacles in the surrounding. Fusion module combines 2D and 3D information to generate a feasible region provided for vehicle navigation.
This paper describes the algorithms for road edge detection in complex traffic situations by an on-board video camera. The information that comes from the motion of a dynamic scene is used to build a new family of fil...
详细信息
This paper describes the algorithms for road edge detection in complex traffic situations by an on-board video camera. The information that comes from the motion of a dynamic scene is used to build a new family of filters called Dynamic Time Filters (DTF). These filters apply morphological opening and closing transformations to the image sequence provided by the camera. This sequence is considered as a 3D image. The DTF-filtered image is segmented by means of the watershed transformation. Two different modes have been implemented for the segmentation. The acquisition mode allows an initial extraction of the current traffic lane. It also produces initial markers which will be used by the second segmentation mode, where an interpolated road/lane model is used to follow and segment the lanes. Some results of the segmentation process are given in complex situations to illustrate the efficiency and robustness of this approach. Information is also given on the real-time performance of the algorithms implemented on a car demonstrator.
暂无评论