Font Size: a A A

Research On Enhancement And Classification Of Mongolian Furniture Patterns Based On L?? Transformation And AGC

Posted on:2021-04-23Degree:MasterType:Thesis
Country:ChinaCandidate:H WangFull Text:PDF
GTID:2381330605473581Subject:Wood science and technology
Abstract/Summary:PDF Full Text Request
At present,In the classification of Mongolian furniture patterns,subjective visual effects are mainly used for classification,and the classification results are not accurate enough.In the process of obtaining Mongolian furniture patterns,due to the influence of historical changes,environmental factors,sampling equipment and other factors in the transmission process,the patterns will be fuzzy and distorted,which will affect the pattern recognition.In order to enhance the fuzzy distortion of furniture patterns,and better improve the recognition rate of patterns,so as to achieve the purpose of accurate classification,this study proposes a Mongolian furniture pattern enhancement algorithm based on Lap transform and AGC(Adaptive Gamma Correction).In this algorithm,RGB(Red Green Blue)color space is transformed into Lap color space,where L(brightness)is the brightness channel,a(Red green dominance channel)is the red green channel,and ?(Yellow blue dominance)is transformed into L?? color space Channel is a yellow blue color channel,which eliminates the mutual interference between RGB color channels.At last,AGC algorithm is used to enhance the output of enhanced furniture patterns.The main contents of the thesis paper include:1.The image enhancement theory and three traditional enhancements of representative DWT-SVD(Discrete Wavelet Transform-Singular Value Decomposition),CLAHE(ContrastLimited Adaptive Histogram Equalization)and gamma correction are introduced.2.This paper proposes a Mongolian furniture pattern enhancement algorithm based on Lap transformation and AGC.The four types of animal,plant,geometric,and text patterns in Mongolian furniture pattern are used as research objects.Correction.enhancement algorithm for comparison,application of peak signal-to-noise ratio PSNR(Peak Signal-to-Noise Ratio),mean square error MSE(Mean Square Error),structural similarity index measurement SSIM(Structure Similarity Index Measurement),information entropy IE(Information Entropy)for objective evaluation.3.Taking animal,plant,geometric,and text patterns in Mongolian furniture patterns as recognition samples,pre-process the patterns using the algorithm and three traditional enhancement algorithms,and then use SVM(Support Vector Machine)radial basis Recognize the enhanced pattern with the angle MSE,brightness MSE,contrast C(Contrast)and homogeneity H(Homogeneity)identification parameters,analyze and compare the identification results.
Keywords/Search Tags:L?? color space, AGC, Pattern enhancement, Pattern recognition, Pattern classification, SVM
PDF Full Text Request
Related items