Research On Extracting Technology Of Stripe Principle Lidar Echo Signal

Comparing with traditional measurement technology,Lidar have many unique advantages.Although the development time is shorter,but Lidar has been rapidly applied to the remote measurement,terrain detection and unmanned obstacle avoidance,etc.Stripe principle airborne laser radar is a typical representative of laser radar.It has occupied the mainstream market of laser radar for its high efficiency and practical characteristics and is playing a more and more important role in both military and committee.Stripe principle laser radar has become a hot research direction now.At present,in the aspect of hardware,stripe principle laser radar has achieved rapid development,but in the aspect of software,it cannot keep pace with the development of hardware.Especially for the early processing part of the original echo signal,the computational complexity and higher difficulty is hard to solve.It has now become an important factor that restricts the development of laser radar.Based on the research background,in this paper,we focus on the extraction technology of the original echo signal for the stripe principle laser radar.The work mainly contains signal denoising processing and information extraction.Firstly,this article expounds the work principle of the stripe principle laser radar,analyses the statistical characteristics of the raw echo signal noise,found that the statistical results conform to the Boltzmann distribution.Then,based on the statistical properties of the signal,we design an edge detection denoising algorithm and a fractal theory information extraction algorithm,the algorithm ideas and calculation process are described in detail.By comparing the denoising effect of traditional mean filtering,median filtering and wiener filtering with that of edge detection denoising method,and introducing relevant evaluation parameters such as SNR,correlation and discrete degree to analyze the denoising results of four ways,we can find that the denoising effect of edge detection denoising method for stripe echo signal is the best.The three main parameters such as the threshold,window size and template characteristics of the edge detection denoising method were studied.After analyzing the influence on denoising effect by changing the three parameters,we found that when the threshold was 10,the window size was 30×30,the template was bar type and the number of non-zero values was 20×30,the denoising effect was the best.Integrated the processing result of the three parameters,it showed that changed in the appropriate range,threshold selection had the largest effect on the denoising performance.Comparing the fractal information extraction algorithm with the common Gaussian fitting,centroid algorithm and Snakes active contour extraction algorithm,we analyzed the characteristics of each algorithm theoretically.In addition,the echo signal information extraction effect of centroid algorithm,Snakes active contour model algorithm and fractal information extraction algorithm was compared and analyzed,we found that the fractal information extraction method had higher extraction efficiency.