Research On Enhancement And Recognize Of Mongolian Furniture Patterns Based On Singular Values And Gamma Functions In Frequency Domain

The Mongolian furniture pattern is not only an inheritance and development of the centuries-old Mongolian culture,but also a unique creation of the Mongolian people and a visual art form that incorporates the spirit and emotion of the nation.Mongolian motifs are diverse and mixed,with heavy lines and magnificent colours,and have been widely used from the time of their formation to the present day.The majority of the information that people obtain comes from visuals,and images,as an important carrier of communication in people's lives,are blurred and distorted due to age and climate change.In this thesis,based on the frequency-domain singular value and the improved frequency-domain singular value method,an enhancement method of the frequencydomain singular value and gamma function is proposed to enhance three types of Mongolian furniture patterns: animal,plant and geometric.The enhanced grain patterns are used as recognition samples using a recognition method based on SVM Gaussian kernel functions using four recognition parameters: uniformity,contrast,entropy and inverse disparity.The main elements of this thesis research are.1.an introduction to intra-frequency domain singular value and an improved intrafrequency domain singular value method with three patterns of Mongolian furniture.2.To propose enhancement methods for singular values and gamma functions in the frequency domain based on singular values in the frequency domain and improved singular values in the frequency domain methods.The smooth wavelet transform(SWT)and discrete wavelet transform(DWT)are used to decompose the grain samples,interpolate and inverse discrete wavelet transform to obtain resolution-enhanced grain samples,then the coefficients ? obtained from the singular values are used to enhance the image contrast,and finally the Mongolian animal,plant and geometric grain samples are corrected and enhanced by the adaptive gamma function.3.The enhancement methods of the in-frequency domain singular value and gamma function are compared with the in-frequency domain singular value and the improved infrequency domain singular value methods,and the enhancement results are subjectively analysed,while the peak signal-to-noise ratio(PSNR),mean square error(MSE),structural Objective evaluation by using singularity in the frequency domain.4.The Mongolian furniture patterns were pre-processed using the frequency-domain singular value,the modified frequency-domain singular value method and the augmented method of frequency-domain singular value and gamma function,respectively,with animal,plant and geometric furniture patterns as recognition samples,and the recognition parameters Entropy,Contrast,Homogeneity,Homo,and Support Vector Machine(SVM)were used for recognition.5.In this thesis,the proposed enhancement method with singular value in frequency domain and gamma function is compared with the singular value in frequency domain method and the improved singular value in frequency domain method,and it is concluded from the results of subjective analysis and objective evaluation index that the enhancement method with singular value in frequency domain and gamma function has better enhancement effect.Using uniformity,contrast,inverse disparity and entropy as recognition parameters,the enhanced Mongolian motifs were recognized using SVM Gaussian kernel function,and the enhancement methods of intra-frequency domain singular value and gamma function had the highest recognition rate for animal,plant and geometric furniture motifs.