Research Of The Algorithm For Evaluating The Cleanliness Of The Road Based On FPGA

Road cleaning is one of the important concerns in the construction of healthy city,and the degree of cleanliness index is an important basis for judging whether the road is clean.If the evaluation of cleanliness of the road is carried out manually,the efficiency is low,and most algorithms for evaluating cleanliness of the road cannot meet the requirements of high real-time performance,low cost and low power of operating vehicles after deploying.How to solve the algorithm for evaluating cleanliness of the road and meet the high requirements of operating vehicles is a problem worth studying.In view of the above problems,this thesis studies the algorithm for evaluating cleanliness of the road and how to deploy the algorithm,mainly including the following work:The thesis study an algorithm for evaluating cleanliness of the road based on FPGA.The algorithm based on machine vision principle,and can detect the characteristics of garbage from mult-channels in parallel.In a variety of working scenes of system,the algorithm can have high evaluation accuracy,and the algorithm on FPGA after deployment can satisfy the sweeper high real-time performance,low cost,low power requirements.By analyzing working scenes of operating vehicles,the project summarized 8 scenes of asphalt and cement road,which includes sunny day,cloudy day,after rain and night,and produced the data sets that includes 800 images for these scenes.In order to solve the problems of light inhomogeneity and complex road in the above scenes,the algorithm in this thesis firstly improves the image quality through image enhancement,and then improves the Nrgb feature space to suppress light interference,so as to improve the accuracy and robustness of the algorithm.Experiments show that the algorithm has a high accuracy rate in a variety of working scenes,the accuracy rate can reach 0.9 in sunny and cloudy scenes,and 0.75 in night and rainy scenes.Finally,the algorithm was deployed on the Xilinx Zynq7020 platform,and its performance was tested to be sufficient to meet the requirements of system.