Study On Measurement Algorithm For Surface Area And Lumen Volume Of Industrial CT Three-dimensional Images Based On C-V Model

With the rapid development of technology and industrial production, the inside and outside shape and performance of object is proposed by many fields. And the demand for performance, appearance and other aspects of product is higher and higher, so that kinds of curved surface are more used in industrial product. The surface's measurement is the base of surface's accuracy check and reverse engineering, to realize the surface's model construction, shape improvement, performance evaluation, and inverse manufacturing, we must obtain 3d parameter information of the surface.So researching the measurement method of curved surface has great significance.The traditional measurement method was divided into the contact and the non-contact measurements based on the measurement style. The contact measurement method has great accuracy, versatility. But the deficiency is that has great computation, high cost and be influenced by the materials of the measured object. The non-contact measurement method could measure some non-contact objects, such as radiator, high temperature object, a relatively rapid speed, has low accuracy compared to contact measurement method. However, the measurement of the internal structure in products was difficult to achieve, especially the closed lumen by traditional measurement methods. The industrial computed tomography is that can detect the object by ray source launching ray from multiple angles, use detector to collect the information which is the ray after decay that through the detected object, use computer and the image reconstruction algorithm to reconstruct the detected parts, then obtain an orderly series of slice images. The external and internal structure size of work piece could be non-destructive detection, especially be suitable for the measurement analysis of closed lumen of work piece.The traditional image method was removing the noises from the image before segmenting, on this paper, taking the measurement of closed lumen of internal woke piece, research a measurement algorithm for surface area and lumen volume of industrial CT three-dimensional images based on C-V model.The main contents in the thesis are as follows:Firstly, the image segmentation methods are analyzed, measuring the ICT image, the precision of image segmentation will directly affect the precision measurement. Therefore, image segmentation is the crucial step to three-dimensional measurement. The C-V geometric active contour model is based on the theory of partial differential equations. The segmentation result which is globally optimal, continuous edge can be gained by using C-V model, and C-V model which has strongly noise immunity is the effective image segmentation method currently. Therefore, we use C-V model to segment the industrial computed tomography images, extract the measured target area, then track the contour points of the target area, and store by chain code; secondly, the center points of outline edge on the surface are acquired, calculate the distance data between points on the surface and the center point. Then use Gaussian filter to smooth contour, so that the edge and burr's effect on measurement can be eliminated; thirdly, establish triangles between contour points of each layer, calculate the surface area by computing the triangles area. Use Green formula to compute the area of each piece, use Platform formula to compute the lumen volume area. The experimental results shows: the segmentation method on this paper can segment objective regions effectively, compared to the traditional measurement methods the method could more accurately acquire curved surface and lumen volume areas.