Road And Object Detection Based On Multilayer Laser Radar

Today there is a huge number of vehicles(cars, trains, planes) in the world. In one hand, it greatly simplifies our daily lives but in the other hand, it poses many problems that must be a solved. Research about vehicle recognition and adaptive cruise control is recently emerging. Many well-known vehicle manufacturers introduce new technologies in their production that gives the opportunity to increase the safety of our lives. Vehicle recognition and adaptive cruise control using various sensory resources. There are many sensors like cameras, GPS, accelerometers, radars etc., that are used for detecting edges of the road, traffic signs, cars or pedestrians. There are many different sensors to access the objects' data information.This paper presents obstacle road detection. To accomplish this task, from the entire set of available sensors the multilayer LIDAR(Light, Detection and Ranging) with four levels was chosen. This allows obtaining reliable information about the location of objects. The so-obtained data have parameters relative to the coordinate system, the distance of each point from the sensor, which opens up great opportunities for development.Primarily, this paper considers how to obtain data from the laser, with detailed consideration of its options. At this stage, the important step is data preprocessing, which includes consideration of only those points that are included in the scan range. The elimination of noise.The second step is the most important task is how to detect the road and objects on it. The solution to this problem requires the use of a variety of methods. It was viewed a large number of sources in this field and the following data processing methods. It is known that for the detection of the road can be considered a method of determining direct. Today, one of the most widely used methods is the Hough method. Its detailed implementation is proposed in this work. For the detection of objects is regarded, the least squares method. This method works with the coordinates of points, which gives the opportunity without any difficulties to use it for our purposes. Based on the results of the Hough method and the method of least squares is used bounding Boxing. This allows us obtain the distance of the object to the data source.As a programming environment was chosen the software Matlab. This software has ample opportunities to work with matrices and their processing. It has a fairly simple interface that greatly simplifies the work. All the result proposed in the draft is obtained programmatically.In order to verify the performance of the algorithm mentioned in this paper, the software of driving assistant system is designed and programed, and are tested together with BJUT-IV. The results show that the system has reached the expected target and has a good environment adaptability and real-time performance.