Research Of Uyghur Handwriting Identification Based On The Edge Stroke Structure

Pattern recognition is one of the basic human intelligence, human computermaster intelligence to help replace part of human mental. With the development of thefield of pattern recognition, computer handwriting identification techniques to becomea hot research in pattern recognition, especially in ethnic minority areas, the writeridentification research is still in its infancy. This highlights the significance of the ofUighur writer identification research.Writer identification research covers the knowledge in the field of imageprocessing, pattern recognition and classification design, you need to solve the typicalproblems in various fields. By analyzing the characteristics of the Uyghur textproposed an edge strokes characteristic structure theory to construct quantitativemodels of handwriting, and design characteristics of the structure extraction algorithm.Mathematical model using a variety of strategies in data classification, data similaritydetection. In this context, the research content Uighur writer identification provides acomplete system solution.This paper studies content:1. Pre-image processing, to assess the experimental feasibility normativewriting requirements specifications and handwriting samples of handwritingsamples and supervise the writing in accordance with the standards ofwriting to complete the writing task, use the scanner scans of200volunteershandwriting ID, establish handwriting library. Secondly, the discussion ofimage denoising, handwriting samples and experiment with the best softthreshold denoising processing model of handwriting original picture. Thenuse a weighted grayscale formula in the the denoising basis of gray scaleprocessing. Again, the use of Otsu binarization of handwriting image. Finally,use a variety of edge operator extraction handwriting edge image, and summarizes the experimental process used herein, the system and imageprocessing layer.2. Pattern recognition processing is the core content of this article writeridentification, the stage of detailed elaboration of the concept and itsclassification of the edge strokes characteristic structure proposed in thispaper, and discusses in detail the implementation method of thecharacteristic structure of the edge strokes. Details edge strokescharacteristic structure extraction algorithm, a detailed analysis of theperformance of the extraction algorithm Finally, the time complexity of themodel.3. Classifier processing stage, the first to accept the classification theory andmathematical models. Secondly, the classification of a variety of strategies,through test eigenvectors data, analysis of the causes of the results, and toidentify and select the design to meet the requirements of the classification.