We propose an edge filter for road white-line detection. Many methods for white-line detection have been proposed for standardized roads. These methods are composed of two stages of processing, i.e., edge detection on an image taken with a camera and then extraction of edge clusters of white-line contours by model fitting. It is difficult to apply these methods to non-standardized roads of for which modeling are difficult. To expand the scope of white-line detection to common roads in the future, it is necessary to achieve processing through clustering without models. However, while clustering can apply to diverse contour lines, there is concern about degrading noise reduction that has so far been done by model fitting. In this study, for the first-stage processing, we developed an edge filter that utilizes the characteristics of white-line contours and detects noise correctly. This filter uses brightness-gradient approximation by discrete values, for which we obtained an idea for a non-linear filter that approximates a low-pass filter plus differential calculus. By applying the method to the images taken by on-board camera, we demonstrate that white-line detection that can apply to diverse road environments but is hardly affected by noise can be realized through combination with model-less clustering.
The ability to estimate selfLposition accurately and robustly is widely regarded as one of the findarnental prcconditions to achieve next − generation car navigation and driver assistance systems . The authors have proposed a localization technology that estimates self − position by matching the three −dirnensional map and entire circumference fish − eye cameras ' images with a particle filter . It has been confrrmed that this technology can estimate self − position with high accuracy and strong robustness ifthe vehicle moves at low speed on a flat road .In order to estimate selfLposition on the slopes and during high − speed moving , the authors improve this technology to be able to set the scattering range of particles dynamically by referring to 出 e 血 g condition d 創 adient infbmlation ofthree − dimensional map .
We propose curve contour detection algorithm for road white line detection based on Helmholtz principle. White line detection is widely used in driver support systems used mainly in highway or major arterial road. As the common road will be the target of operational area for autonomous vehicle, it is thought to be necessary to develop a new detection algorithm that can deal with various types of road. This paper proposes model-less algorithm that is constructed on a new edge feature inspired by Helmholtz principle through the analysis of the limit of Hough transform. This feature is basically same as Hough defined feature of edge count on the line except two remarkable points. The one is the restriction of count area and the other is the way of count which affords to detect curve line as well as straight line. Implementation by convolutional neural network is explained and the relation between tunable parameters and the detection performance as well as the processing time are discussed. Comparison between conventional methods such as Hough transform or machine learned contour detection algorithm BEL is explained for test image and images taken by on-board camera to show the superiority of proposed algorithm. We demonstrate that proposed algorithm that can apply to diverse road environments but is hardly affected by noise can be realized.
We propose image processing method for road white-line detection using brightness gradient direction of edges. Many methods for white-line detection have been proposed for standardized roads. These methods use model fitting for the detection of edges on white-line contour. So it is difficult to apply these methods to non-standardized roads for which modeling is difficult. To expand the scope of white-line detection to common roads in the future, it is necessary to achieve processing without models. Clustering based on position proximity of edges is one approach. However, there is concern about degrading noise reduction that has so far been done by model fitting. In this study, we developed an edge clustering method that utilizes the characteristics of edges on white-line contours; proximity of position as well as proximity of brightness gradient direction. In the proposed method, edges are first clustered based on proximity of gradient direction. And for each cluster, edges are again clustered based on proximity of position. Edge filter bank is specially designed for the first clustering and the effectiveness compared to conventional filter bank is explained. By applying the method to the images taken by on-board camera, we demonstrate that white-line detection that can apply to diverse road environments but is hardly affected by noise can be realized.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.