Research On Industrial Application Of Close Range Digital Photogrammetry

With 150-year history, today photogrammetry has made fundamental changes ascomputer and digital camera technology develops. The evolution of photogrammetryhas gone through three phases including analog, analytical and digital. Initially it foundits application in map making and acquisition of geographical information in aerial andaerospace photogrammetry. Currently due to the great price drop-down of computerhardware and digital camera, its application has extended to industrial fields, measuringproducts and parts. Relative to far-range aerial and aerospace photogrammetry,close-range photogrammetry came into being. As the technology gets improved andmature with the wide application, and considering its ever-growing application futureand high-tech nature, we selected this subject as the thesis project.Photogrammetry is a branch of measurement technology; it computes threedimensional scene coordinates of some feature points based on sets of photograph pixelcoordinates, with the objective of recovering depth information through triangulation.Triangulation means taking pictures of the same scene from two or more differentcamera positions, and making use of the parallax to recover the depth information. Inorder to triangulate the measured points, we must know the orientation of the pictures.However, in order to orient the pictures, we must know the coordinates of the measuredpoints, it appears there is a contradiction. To solve it here using the secret weapon (orkey technology)—the bundle adjustment, it has the capability to figure them both outsimultaneously. The essence of the technique is to use the nonlinear least squareoptimization method to optimize the simultaneous equations so that the mean squareerror between the estimated mean and computed value of each photograph is minimized.To ensure the smooth implementation by this method, the corresponding points indifferent images must be automatically matched. To facilitate matching and extraction,feature points of the scene object must be salient; this can be realized by putting aretro-reflective marker as explained in the article. This paper also elaborates thephotogrammetry planning processing; put emphasis on the analysis of factors affectingmeasurement accuracy and precautions to ensure a high accurate measurement.Our work involved in this project mainly is the following:1. Did research on photogrammetry history and measurement principle;2. Studied photogrammetry process and high accurate measurement precautions;3. Studied mathematical theory related to photogrammetry including: projectivegeometry, optimization algorithm, linear regression, methods of solving for projectivematrix, essential matrix, and fundamental matrix, extraction of internal and externalparameters of camera, and the 3D reconstruction method using projective matrix. 4. Using Matlab programs to validate the above theory and algorithm, includingmanually pick up matching points, corner extraction, camera calibration using twodifferent group of pattern orientations, compute camera internal parameter, displayspatial relationship between camera and pattern, finally use Delaunay triangulation andtessellation to reconstruct two and three dimensional scenes.