Research Of Cashmere Feature Parameter Extraction Techniques Based On Image Analysis

At present, the study of feature extraction of fiber has become a hot field of the textile industry and image processing. This paper mainly studies the cashmere fiber feature extraction techniques based on image analysis, including studying the image processing program of cashmere fiber and studying the measurement method of the feature parameter of cashmere fiber.The image processing object is cashmere fiber image obtained under the optical microscope, through the requirements analysis, outline design, and analyzing the character of cashmere fibers images and requirements of extracting information,then preliminary confirms the image processing function modules , such as image de-noising, image enhancement, image segmentation and modification treatment. As the images collected under optical microscope have limits of edge blur, gray value does not differ greatly, boundary constraints of piecemeal, in order to get continuous fiber scale edge information, study each module's achievement methods, and then finally find a more suitable processing method for this paper. Therefore, this paper uses Matlab7.0 to do simulation experiments. A large number of image processing methods are tested and compared and chooses a better method that obtained better processing results, and to do simulation experiments using a variety of treatment methods combined. Verified through simulation experiments, this paper finally brings forward image processing programs that meet the needs of this paper. The image processing programs are the image gray-oriented, using Laplacian to reinforce the edge of fibers and scales, using median filter to remove noise, using Canny operators to edge detection edge, and using morphological expansion and refinement algorithms to do final processing. Using the processing programs to process the images of this paper can get the binary image whose fiber edge and scale edge contains continuous and single pixel information.After using the image processing program to process the image, this paper studies the the cashmere fiber characteristic parameters extraction and uses appropriate methods to achieve the characteristic parameters extraction. This paper studies the indexes measurement such as triangle method, central axis algorithm used to measure the fiber diameter, then analyzes the lack of common indexes measurement. This paper designs the sub-measurement algorithm to achieve the cashmere fiber diameter measurements, designs the central axis measurement algorithm to achieve the scale height measurements, greatly reducing the impact of human factors and improving the measurement efficiency. Finally, based on MFC application framework, this paper uses object-oriented programming ideas to realize the automatic extraction of fiber diameter, scale length, scale density and the Ratio of diameter and scale length, and through the ADO to access SQL Server database, then import the data into the database to facilitate the use of follow-up work.