Demensional measurement of 3-D objects through stereometric image processing

Abstract

The purpose of this thesis is to develop a method of measuring the dimensions of 3-D objects by using a pair of T.V. cameras. The major areas of this algorithm is based upon the fact that a pair of images do not exist on the identical plane. If views of the same object from two different positions are available, we can measure the 3-D coordinate through the concept of correlation between gray levels of a pair of images and the concept of triangulation. The procedure can be summarized as follows; Firstly, the calibration matrices between world coordinate and each image coordinate are obtained, then a point is given in one image. Secondly, the inverse perspective line equation on one image plane corresponded a point in the other image plane is calculated and correlation techniques are used to find a maximally matched point pixel in the second image. Finally, the 3-D information is obtained through the concept of triangulation. And this study is briefly mensioned that two images exist on an identical plane. The system is operated and provided three-dimensional measurement error of 2\%.