Research On Detection Methods Of Cotton Fiber Based On DSP

In recent years,with the rapid development of industrial automation technology, using the automatic detection and control equipment to replace high-intensity, high-density human operator is becoming a major trend. The detecting machines of foreign fibers in our country have lower efficiency of recognition, and often leave out much foreign fibers; The price of foreign fibers sorting equipments made by foreign equipment manufacturers are too high to be accepted by our cotton textile Enterprises, and the foreign fiber sorting equipment from abroad can not support the use of the existing complete set of equipment Under the background,the subject comes into being so as to develop our foreign fibers detecting device with our intellectual property, get rid of dependence on foreign products,save expenditures of industry greatly and raise our level of automation in the textile industry to increase the competitiveness of cotton spinning enterprises. This system is composed of DSP to realize high-speed and real-time detection of foreign fibers in cotton.TMS320C6000 DSP chip is produced by TI which is a dedicated chip for image processing.Foreign fibers detection algorithm is the key part in this paper.The preceding work in the paper is mainly on researching and analysising the knowledge of digital image processing and pattern recognition, image pre-processing part of the raw cotton mainly takes advantage of the knowledge of mathematical morphology processing.The following testing part advances "three-color RGB threshold recognition algorithm", which means this algorithm can recognize foreign fiber with great difference in color but not with similar color.For this problem,based on multi-band spectral irradiation of fiber color analogy, this paper make use of image fusion method,which can better detect the foreign fibers with similar color.In the latter half of the paper,the foreign fiber detecting device hardware has been studied and discussed.This detection system consists of five modules: cotton feeding module, optical collecting module, image collecting module, foreign fibers detecting module and foreign fiber automatic removing module.At last, based on all the researchs, foreign fiber detecting recognizing algorithm is validated according to simulation.Experimental results show that the proposed research programs and detection algorithm are effective to achieve the desired requirements and feasible for the future in-depth study.