Research On The Key Technologies Of Navigation For Quadrotor UAV Based On Laser Radar

In today's information society,video image has gradually become the main carrier for people to obtain information,so the quality of video image has attracted more and more attention.There are many problems affecting the image,some of them are hardware problems,but the most important one is the change of environment.There are many factors that affect the environment,such as haze,rain and so on.This thesis focuses on the research of video images in different lighting environments.The design algorithm can automatically identify the lighting environment of the image,and use appropriate image enhancement methods for image processing.Then,according to this algorithm,a prototype system of the monitoring system is designed and built to verify its practicability.This thesis first introduces the design of the calculation method of image eigenvalue without reference,explores the performance of the same object and background eigenvalue under different lighting conditions,and tests the image processing ability of different enhancement methods,and carries out the horizontal contrast experiment under the control of lighting as a single variable.Then,the results are summarized statistically,and the application scope and selection criteria of each enhancement algorithm are established.Based on this,the image quality detection and classification algorithm is designed.In this thesis,aiming at the problems of the related algorithms based on Retinex theory in image enhancement,an improved algorithm of image edge enhancement algorithm based on Retinex theory is proposed.Firstly,for the problem that the edge of detail is not obvious in the enhancement result,the Laplacian algorithm is proposed to further process the detail image.At the same time,for the problem that the image is prone to halo after being enhanced,the gamma correction method for the illumination image is proposed.Through the experimental comparison,the algorithm has better reinforcement effect and better adaptability.Then,based on the requirements of real-time and delay in the actual operation of the monitoring system,a multi-threaded strengthening method is designed.Then,the similarity between adjacent frames is studied and demonstrated,and the gray histogram correlation algorithm is used to make quantitative data requirements for this problem.Based on this,the judgment and processing algorithm for the similarity of adjacent frames is designed.Finally,this thesis synthesizes the previous research and algorithm,designs a video image quality enhancement module for the video monitoring system,and builds a monitoring prototype system around it.Based on this,the field experiment is carried out,and the running results are given to verify the effectiveness of the system and the feasibility of the algorithm.