Font Size: a A A

Research On Three-dimensional Reconstruction Algorithm Based On Computer Vision

Posted on:2016-08-23Degree:MasterType:Thesis
Country:ChinaCandidate:X LiuFull Text:PDF
GTID:2428330542989514Subject:Control engineering
Abstract/Summary:PDF Full Text Request
In the production process of logistics information management,it is a vital part to measure the volume of material stack effectively.In order to improve the automation standard for materials measurement,it was researched that the three-dimensional reconstruction algorithm based on computer vision with fast speed and high accuracy.The main contents of this thesis are listed as follows:(1)The scatted points were preprocessed.Since the annular working platform is not parallel to the material stack platform,the three-dimensional data begins to deflect.The Z coordinates of boundary points are reduced to near zero.To calculate the inclination angle,the boundary dates should be processed.All the surface points of material stack were collected completely,which leads to higher accuracy than ever.(2)The theories of three-dimensional construction and volume calculation were analysed in this thesis.All sorts of spatial interpolation algorithm were contrasted to obtain the advantages and disadvantages.The Kriging interpolation algorithm which is suitable for the experiment is adopted and improved.At first,the known sampling point data is dealt with to obtain the experimental variation values.Then,the experimental variation values are fitting through theoretical variation model to acquire the variation function model.Finally,the sampling points are substituted to Kriging equation groups.The sampling point values can be calculated.(3)Considering the spatial correlation among the sampling points,the sampling point search strategy was improved.The local search method is replaced by the whole.The direction search method was designed in this thesis.There is high accuracy and fast speed in this method.(4)The experiment adopted the whole interpolation Kriging interpolation and local Kriging interpolation to three-dimensional construct and calculate the material stack volume.The purpose is to compare the accuracy of these methods.The result is that adopting the local Kriging interpolation algorithm can be closer to real surface of material stack and the accuracy is higher than others.To some extent,this research realizes the automatic measurement of large complex shape material stacks of large and complex shapes.Modeling and testing in the laboratory to verify the feasibility and accuracy of this program.The method is simple,low cost,powerful generalization.
Keywords/Search Tags:Computer Vision, Camera Calibration, Material Stack, Spatial Interpolation, Three-dimensional Reconstruction
PDF Full Text Request
Related items