Road Detection Based On Multi-frame Video Image

The key for vision-based road detection is the ability to accurately determine road image pixels as the road or not.Vision-based road detection is very challenging for heavy traffic flow and continuously changing background.Monocular vision has low cost,whereas visual information is difficult to be processed,thus road detection algorithm based on monocular vision has become a research hotspot.In this paper,for the specific application of unmanned ground vehicles,we study how to detect road based on monocular vision in noisy road environment and the proposed algorithm for road detection provides a high accuracy and robustness.The main work and research results are as follows:In order to improve the speed of road detection and reduce the interference of irrelevant information,a method of setting the r egion of interest(ROI),which integrates the driving parameters of the unmanned vehicle with multi-frame video image,is put forward.According to the driving parameters of the unmanned vehicle,ROI is a part of the whole image which is determined using a calibrated camera.The detection result is seriously affected by the quality of image.An algorithm of denoising and deblurring using directional filters is improved.It makes the algorithm satisfy the real-time requirement.Applying a directional low-pass filter to the input image greatly reduces the noise level,while preserving the blur information in the orthogonal direction to the filter.This approach reduces the noise without degrading blur information,thereby producing better kernels.Meanwhile,it generates a high-quality result with a given estimated kernel by two-step iterative optimization.In addition,the combination of multi-frame video images road,on the basis of a frame noise to blur the image,combined with noise reduction feature differences between adjacent frames to blur,noise reduction can be derived blurred image of the current frame to reduce time complexity,to meet real-time.A novel algorithm for road detection based on superpixels and anisotropic heat diffusionis proposed to improve the detection robustness.Firstly,the method oversegments the enhanced image into small homogeneous regions which are called superpixels with the superpixel segmentation method.Then,we cluster superpixels into several region blocks according to the thermodynamic energy diffusion theory.Finally,the algorithm extracts the road region combining the prior knowledge with the largest and most coherent principle of cosegmentation.Based on multi-frame video image and the detection of the current frame road area,in front of the adjacent frames may be combined,synergistic combination of segmentation theory,can quickly get the road region of the current frame.Implement denoising and deblurring algorithm and road detection algorithm based on superpixels and anisotropic heat diffusion within ROI.Experimental results demonstrate that it is robust to lighting variations and heavy traffic.It provides the better performance when compared with several state-of-the-art methods.