Three-Dimensional Reconstruction and Evaluation Based on CT Images

Abstract

Two-dimensional computed tomography (CT) images are not enough for further analysis. Only the points with the same Z coordinates can be measured on the two-dimensional images, and the points on the different layers can not be measured. In order to promote the development of the industrial CT, the three-dimensional reconstruction and analysis system is researched in this paper. The three-dimensional reconstruction system is based on the sequence of the two-dimensional images. Firstly, the edges acquired after threshold segmentation should be found in each image. Secondly, the corresponding edges describing the same surface should be found in the different images. Thirdly, the corresponding points on the corresponding edges should be found according to their similarity. After all the points are found, the contours in the different images can be connected. The three-dimensional reconstruction system provides move, rotation, and zoom tools, so it can give more vivid view for the operators. It seems as if the real object was shown in the computer. The analysis system is based on the three-dimensional reconstruction system; it includes some useful functions, such as distance measurement, angle measurement, volume measurement, cut and installation of a part, peeling off layer after layer, etc. Any points on the images can be measured. All these functions can help the operators to study the object more conveniently. After 50 times measurement of two selected points, the average measurement precision can reach to 0.0007 mm. From a lot of experiments, the three-dimensional CT system is proved that it is more vivid, reliable, convenient, and useful. And it can be extended to any non-destructive tomography technology, such as optical coherence tomography (OCT), magnetic resonance imaging (MRI), single photoemissions computed tomography (SPECT), positron emission tomography (PET), etc.