Research On Data Processing And Visualization Of High-speed 3D-imaging Laser Radar

When we reconstruct surface use 3D-imaging laser radar data, because of there are noises in it, we need to filter the data. This paper simulates the scan process of ladar and use it scan some objects whose characters we has known and gets the simulated data of these objects, by analyzing these data establishes the conception of objects frequency. By analyzing the ladar data of objects that we are familiar with, we get the frequency composition feature of common objects. This paper designs a FIR low pass digital filter in Hamming Window method to filter every row and column of the ladar data and chooses the appropriate parameters through theoretically and experimental research. Research shows that this designed filter can get well effects not only to the data of flat objects but also to the data of some shape rapidly changed objects .Compare with filters we usually use, this filter can bring fewer error and has lesser computation amount.There are two forms of ladar data, one is one dimensional, and another is three dimensional. Correspondingly, there are there are two methods to interpolate the ladar data, one is one dimensional interpolation which bases on the original row and column ladar data, the other is surface interpolation which bases on the three dimensional surface data of objects. This paper analyses several usually used one dimensional interpolation methods, compares their computation amounts an interpolation effects, and concludes that we should take one dimensional piecewise linear interpolation method for ladar data. This paper also researches a surface interpolation method which has fewer computation amount and better effects among the surface interpolation methods.This paper writes imaging software for imaging ladar in OpenGL and VC++. The software can fulfill the 3D reconstruction of ladar data and can satisfy the requirement of real-time imaging.