The Implementation Of Image Separating Algorithm For Pulp Fiber

As the third largest manufacturing industry of the world, and with the combination of the digital image process technique, paper manufacture has obtained widespread focus and research. The height, width, curve characteristic and distributing characteristic of pulp fiber are the important parameters in the pulp industry (Juhua Wang, 1998). The physical parameters of pulp fiber make a great influence on the final paper performance, and effective measurement of pulp fiber characteristics will have a direct relationship with the quality of the final paper. So it becomes more and more important to obtain the main parameters of pulp fiber accurately for the high quality paper manufacturing.Traditionally, the measurement for the shape of pulp fiber will be influenced by artificial factors upon the sampling equipment and data analyzing in a great depth. So measuring the pulp fiber effectively is still an urgent problem needed to be solved.In order to measure different physical parameters of the pulp fiber accurately, we must pick out the geometrical characteristic, of which the most challenging task is to separate and recognize the crossed pulp physical. As it is so common to encounter the overlapping pictures of the paper pulp we sampled in reality which is in other word of the circumstances with multiple pieces pulp fiber crossed together. This will bring difficulties on the subsequent analyzing process. This paper has designed an effective algorithm to solve this problem. Multiple pieces of pulp fiber that are crossed together can be separated into complete pulp fiber.This paper farther discussed the recognition method of generic linear object, and in combination with morphological characteristics of pulp fiber, an effective algorithm is proposed for the subject. First of all we obtain the actual pulp fiber images, then design a high precision pulp fiber image process algorithm and propose a proven separating algorithm base on the theory of image binarization and image analyzing in a traditional method. Besides, the arithmetic has been implemented by programming under the machine vision software -- Halcon, and Microsoft development tools -- Visual C++. NET, finally effectively implement the separation of pulp fiber that is crossed and accurate measurement the characteristic parameters of the pulp fiber.