This work presents a circle detection algorithm for the purpose of detecting roundabout central islands in single-layer laser scans. The algorithm is robust against measurement errors and imperfections of the central ...
详细信息
ISBN:
(纸本)9781479960781
This work presents a circle detection algorithm for the purpose of detecting roundabout central islands in single-layer laser scans. The algorithm is robust against measurement errors and imperfections of the central island's circular form and it can be used for laser scanners with only a few vertical layers. Main features of the algorithm are the semi-convex segmentation and curvature-based segmentation steps and the use of an a priori digital map to classify the detected circles. Experimental results show that, based on its performance, the method can be integrated in a real-time online roundabout perception system for highly automated vehicles.
Finding the center point of the hedge is the key to automated pruning. In a complex outdoor environment, since the horizontal cross-section of the hedge is not a regular circle, and has many approximate circular conto...
详细信息
Finding the center point of the hedge is the key to automated pruning. In a complex outdoor environment, since the horizontal cross-section of the hedge is not a regular circle, and has many approximate circular contours, the performance of mainstream circle detection algorithms is not ideal. To this end, we propose an adaptive hedge horizontal cross-section center detectionalgorithm, named AdaHC, which can obtain the horizontal cross-section center coordinates of the hedge in real time by inputting the top view image of the hedge. The experimental results show that our proposed algorithm is significantly better than other circle detection algorithms. Its recognition accuracy can reach 100%, the average accuracy is over 80% (corresponding to an accuracy of 1 cm), and the average time consumption is 11.6 ms, be robust to the variations in light, hedge color, and background, fully meets the industrial requirements, and lays a good foundation for automatic hedge trimming.
In order to drive through a roundabout an automated vehicle needs to know the roundabout's geometry relative to the vehicle itself. In this work we present a model describing the geometry of a roundabout and a per...
详细信息
ISBN:
(纸本)9781467365963
In order to drive through a roundabout an automated vehicle needs to know the roundabout's geometry relative to the vehicle itself. In this work we present a model describing the geometry of a roundabout and a perception system that computes the parameters of this model relative to the ego vehicle before the ego vehicle enters the roundabout. Only a part of the model's parameters are estimated using on-board sensor observations. The complete roundabout geometry is computed by aligning a detailed geometrical description from an a priori digital map with the estimated parameters. For this purpose a regular navigation map is extended with geometric information about roundabouts. This extended map data is not georeferenced and the presented approach does not require high accuracy global localization or any additional landmarks to be stored in the map. The system's performance has been evaluated based on ground truth data.
Because of many advantages, such as single-sided riveting and no welding, blind rivet nuts are widely used to fasten various metal sheets and pipes. However, fixing the blind rivet nut to the base material is mainly a...
详细信息
Because of many advantages, such as single-sided riveting and no welding, blind rivet nuts are widely used to fasten various metal sheets and pipes. However, fixing the blind rivet nut to the base material is mainly accomplished by manual riveting and robot riveting based on teaching-playback. In order to realize automatic riveting without teaching, this paper proposes a robot automatic riveting method based on machine vision. Firstly, with the help of the light source that increases the difference between the foreground and the background, a binarization method based on morphology is designed to remove lots of useless information in the image. Then, an improved EDcircles algorithm based on the geometric properties of circular arcs proposed to detect circles in the binary image, which also realizes the detection of the head circles of the nut. Finally, in order to obtain the precise position of the nut, an optimization method for the inner circle of the nut is designed. Experimental results show that this method has a high nut detection rate, good positioning accuracy and efficiency, which indicates that it can be used in practical applications.
暂无评论