Research On Detecting Method For Pulp Fibre And Paper Sheet Quality Based On Image Analysis Technology

Based on the image analysis examining the paper pulp fiber and the sheet surface quality is one of the key technology in pulp and papermaking industry, which affects the paper page quality directly, and is becoming an interested topic.Regarding the pulp fiber image analysis, the purpose is to analyze shape and dimension of pulp fiber through the micro image. The relatively uncertainty of fiber shape leads overlap cross of its edges, so the shape should be identified to improve allocation and extraction of the edges. The homogeneity of the sheet is analyzed by image transform, i.e. the spatial domain image is transformed to frequency domain or wavelet domain to obtain the features. The character of the normal or defective construction is extracted by texture analysis to find surface defect of the sheet. The kind of the sheet defect is recognized by using artificial neural network and statistic pattern recognition.The mainly research works of this thesis are as following.1. In this paper, an improved operator was adopted to fiber edge detection, it combined the characteristics of linear objects to improve edge detection scheme using directional wavelet transform. This method is reliable in dealing with complex image edge detection with overlapping structures or branches.On the base of edge detection based on directional wavelet transform, this paper analyzed the simultaneous detection of both borders. Combining the characteristics and spatial distribution of the borders, this paper proposed to use proper cost function to enhance the result of the edge extraction .Regarding the pulp fiber recognition, the method was proposed by using edge and skeleton features to finish the recognition of complex fiber images. which adopted simultaneous border detection to detect edges and direction-changed tracing method to extract skeleton.2. Based on the analysis of several denotation methods that exist in the world, we have combined the wavelet transform multi-resolution analysis technology with the denoting of paper uniform, found a new denotation index--wavelet index (WIn) and method. The results showed that the new algorithm is effective.3. According to the characteristics of paper surface image, we proposed defect inspection scheme witch was based on image fusion, stationary wavelet transform and...