Despite Gabor filtering has emerged as one of the leading techniques for texture classification, a unifying approach to its adoption has not emerged yet. As it is true for Gabor filter bank, the design of a filter bank consists of the selection of a proper set of values for the filter parameters. In this paper, it is intended to find a set of Gabor filter bank parameters optimized for the performance of texture classification system. The application method is suggested to compute Gabor filter parameters based on Genetic Algorithm (GA). The parameters are optimized according to each group of textures. We tested the proposed method with several texture images using a standard database. The experimental results demonstrate the effectiveness of proposed approach as the overall success is about 97.5%.
{"title":"Gabor Filter Parameters Optimization for Texture Classification Based on Genetic Algorithm","authors":"Mehrnaz Afshang, M. Helfroush, Azardokht Zahernia","doi":"10.1109/ICMV.2009.50","DOIUrl":"https://doi.org/10.1109/ICMV.2009.50","url":null,"abstract":"Despite Gabor filtering has emerged as one of the leading techniques for texture classification, a unifying approach to its adoption has not emerged yet. As it is true for Gabor filter bank, the design of a filter bank consists of the selection of a proper set of values for the filter parameters. In this paper, it is intended to find a set of Gabor filter bank parameters optimized for the performance of texture classification system. The application method is suggested to compute Gabor filter parameters based on Genetic Algorithm (GA). The parameters are optimized according to each group of textures. We tested the proposed method with several texture images using a standard database. The experimental results demonstrate the effectiveness of proposed approach as the overall success is about 97.5%.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"09 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125768126","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Support Vector Machine (SVM) is a machine learning algorithm, which has been used recently for classification of hyperspectral images. SVM uses various kernel functions like RBF and polynomial to map the data into higher dimensional space to improve data separability. New kernel functions are used in this paper to classify hyperspectral images which are based on wavelet functions as named Wavelet-kernels. The experimental results indicate that Wavelet-kernels provide better classification accuracy than previous kernels.
{"title":"Using Wavelet Support Vector Machine for Classification of Hyperspectral Images","authors":"Mohammad Hossein Banki, A. Shirazi","doi":"10.1109/ICMV.2009.64","DOIUrl":"https://doi.org/10.1109/ICMV.2009.64","url":null,"abstract":"Support Vector Machine (SVM) is a machine learning algorithm, which has been used recently for classification of hyperspectral images. SVM uses various kernel functions like RBF and polynomial to map the data into higher dimensional space to improve data separability. New kernel functions are used in this paper to classify hyperspectral images which are based on wavelet functions as named Wavelet-kernels. The experimental results indicate that Wavelet-kernels provide better classification accuracy than previous kernels.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121914632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we propose an approach to learn the kernel which uses transferred knowledge from unlabeled data to cope with situations where training examples are scarce. In our approach, unlabeled data has been used to construct an optimized kernel that better generalizes on the target dataset. For the proposed kernel learning algorithm, Fisher Discriminant Analysis (FDA) is used in conjunction with Maximum Mean Discrepancy (MMD) test of statistics to optimize a base kernel using labeled and unlabeled data. Thereafter, the constructed kernel from both labeled and unlabeled datasets is used in SVM to evaluate the results which proved to increase prediction accuracy.
{"title":"Optimizing Kernel Functions Using Transfer Learning from Unlabeled Data","authors":"M. Abbasnejad, D. Ramachandram, R. Mandava","doi":"10.1109/ICMV.2009.10","DOIUrl":"https://doi.org/10.1109/ICMV.2009.10","url":null,"abstract":"In this paper, we propose an approach to learn the kernel which uses transferred knowledge from unlabeled data to cope with situations where training examples are scarce. In our approach, unlabeled data has been used to construct an optimized kernel that better generalizes on the target dataset. For the proposed kernel learning algorithm, Fisher Discriminant Analysis (FDA) is used in conjunction with Maximum Mean Discrepancy (MMD) test of statistics to optimize a base kernel using labeled and unlabeled data. Thereafter, the constructed kernel from both labeled and unlabeled datasets is used in SVM to evaluate the results which proved to increase prediction accuracy.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124092601","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cloud computing is a catchphrase that is flipped around a lot these days to describe the direction in which information road and rail network seems to be stirring. The concept, is that immense computing data will reside someplace out there in the anonymous place (in spite of the computer space) and we'll bond to them and utilize them as needed. This research paper presents basic issues regarding data usage and processing in cloud computing and their limitations. An attempt to propose appropriate solutions for these underlying issues has also been made.
{"title":"Data Processing Issues in Cloud Computing","authors":"A. Khalid, H. Mujtaba","doi":"10.1109/ICMV.2009.31","DOIUrl":"https://doi.org/10.1109/ICMV.2009.31","url":null,"abstract":"Cloud computing is a catchphrase that is flipped around a lot these days to describe the direction in which information road and rail network seems to be stirring. The concept, is that immense computing data will reside someplace out there in the anonymous place (in spite of the computer space) and we'll bond to them and utilize them as needed. This research paper presents basic issues regarding data usage and processing in cloud computing and their limitations. An attempt to propose appropriate solutions for these underlying issues has also been made.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"133 8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131298327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
There are many factors affect the warship Life Cycle Cost (LCC), the importance of every factor is different, and the relationships between factors are correlated. In order to establish the precise LCC model, the Principal Components Regression (PCR) and Partial Least Squares Regression (PLSR) are proposed to reduce the correlativity between factors which affect the modeling of LCC. However, the components often don’t strongly explain the dependent variables when filtering principal components in the independent variables. Therefore, the improved PCR with Rough Set is proposed to overcome the correlativity between the variables, which could choose the important parameters and reduce the unimportant parameters in the modeling of LCC. The modeling of the process and the regression model are described in the content. Compared with the method of PCR and PLSR, the precision of the improved PCR with Rough Set is much higher.
{"title":"Improved Principal Components Regression with Rough Set and its Application in the Modeling of Warship LCC","authors":"Xiao-Hai Zhang, Jia-shan Jin, Jun-bao Geng","doi":"10.1109/ICMV.2009.25","DOIUrl":"https://doi.org/10.1109/ICMV.2009.25","url":null,"abstract":"There are many factors affect the warship Life Cycle Cost (LCC), the importance of every factor is different, and the relationships between factors are correlated. In order to establish the precise LCC model, the Principal Components Regression (PCR) and Partial Least Squares Regression (PLSR) are proposed to reduce the correlativity between factors which affect the modeling of LCC. However, the components often don’t strongly explain the dependent variables when filtering principal components in the independent variables. Therefore, the improved PCR with Rough Set is proposed to overcome the correlativity between the variables, which could choose the important parameters and reduce the unimportant parameters in the modeling of LCC. The modeling of the process and the regression model are described in the content. Compared with the method of PCR and PLSR, the precision of the improved PCR with Rough Set is much higher.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122034106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Quality control is very important in the iron and steel industry to ensure that products meet customer requirements. Flatness is one of the most important features of rolled products, and it is used to estimate the final quality of the resulting product. Therefore, flatness control, which requires precise flatness measurements, is of vital importance during rolling. This work proposes a machine vision system for flatness measurement based on the projection of a laser stripe over the surface of the steel strip. The flatness measured cannot be used as easily as the feedback of the flatness control system due to the huge amount of information it contains. In order to solve this problem, a feature extraction method based on Legendre polynomial fit is also proposed.
{"title":"Machine Vision System for Flatness Control Feedback","authors":"R. Usamentiaga, J. Molleda, D. García, F. Bulnes","doi":"10.1109/ICMV.2009.14","DOIUrl":"https://doi.org/10.1109/ICMV.2009.14","url":null,"abstract":"Quality control is very important in the iron and steel industry to ensure that products meet customer requirements. Flatness is one of the most important features of rolled products, and it is used to estimate the final quality of the resulting product. Therefore, flatness control, which requires precise flatness measurements, is of vital importance during rolling. This work proposes a machine vision system for flatness measurement based on the projection of a laser stripe over the surface of the steel strip. The flatness measured cannot be used as easily as the feedback of the flatness control system due to the huge amount of information it contains. In order to solve this problem, a feature extraction method based on Legendre polynomial fit is also proposed.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"146 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126188866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Watermarking is a technique which allows an individual to add hidden copyright notices or other verification messages to digital audio, video, or image signals and documents. In our proposal, for Audio Watermarking, a Watermark is encrypted using RSA Algorithm and is embedded on the audio file using LSB technique. LSB technique is an old technique which is not very robust against attacks. Here, in audio watermarking we have embedded the encrypted watermark on the audio file, due to which removal of the watermark becomes least probable. This would give the technique a very high robustness. In the retrieval, the embedded watermark is retrieved and then decrypted. This method combines the robustness of Transform domain and simplicity of spatial domain methods. For image Watermarking, DWT technique is used. DWT technique is used in Image watermarking. Here, we have embedded the watermark in the image as a pseudo-noise sequence. This gives a remarkable security to the image file as only if the exact watermark is known can the embedded watermark be removed from the watermarked image.
{"title":"Effective Watermarking of Digital Audio and Image Using Matlab Technique","authors":"S. Subbarayan, S. K. Ramanathan","doi":"10.1109/ICMV.2009.33","DOIUrl":"https://doi.org/10.1109/ICMV.2009.33","url":null,"abstract":"Watermarking is a technique which allows an individual to add hidden copyright notices or other verification messages to digital audio, video, or image signals and documents. In our proposal, for Audio Watermarking, a Watermark is encrypted using RSA Algorithm and is embedded on the audio file using LSB technique. LSB technique is an old technique which is not very robust against attacks. Here, in audio watermarking we have embedded the encrypted watermark on the audio file, due to which removal of the watermark becomes least probable. This would give the technique a very high robustness. In the retrieval, the embedded watermark is retrieved and then decrypted. This method combines the robustness of Transform domain and simplicity of spatial domain methods. For image Watermarking, DWT technique is used. DWT technique is used in Image watermarking. Here, we have embedded the watermark in the image as a pseudo-noise sequence. This gives a remarkable security to the image file as only if the exact watermark is known can the embedded watermark be removed from the watermarked image.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114729190","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Lei Yunqi, Dongjie Chen, Meiling Yuan, Qingmin Li, Zhenxiang Shi
An approach of 3D face recognition by using of facial surface classification image and PCA is presented. In the step of pre-processing, the scattered 3D points of a facial surface are normalized by surface fitting algorithm using multilevel B-splines approximation. Then, partial-ICP method is utilized to adjust 3D face model to be in the right front pose for a better recognition performance. By using the normalized facial depth image been acquired through the two previous steps, and by calculating the Gaussian and mean curvatures at each point, the surface types are classified and the classification result is used to mark different kinds of area on the facial depth image by 8 gray-levels. This achieved gray image is named as Surface Classification Image (SCI) and the SCI now represents the 3D features of the face and then it is input to the process of PCA to obtain the SCI eigenfaces to recognize the face. In the experiments conducted on 3D Facial database ZJU-3DFED of Zhejiang University, we obtained the rank-1 identification score of 94.5%, which outperformed the result of using PCA method directly on the face depth image (instead of SCI) by 16.5%.
{"title":"3D Face Recognition by Surface Classification Image and PCA","authors":"Lei Yunqi, Dongjie Chen, Meiling Yuan, Qingmin Li, Zhenxiang Shi","doi":"10.1109/ICMV.2009.61","DOIUrl":"https://doi.org/10.1109/ICMV.2009.61","url":null,"abstract":"An approach of 3D face recognition by using of facial surface classification image and PCA is presented. In the step of pre-processing, the scattered 3D points of a facial surface are normalized by surface fitting algorithm using multilevel B-splines approximation. Then, partial-ICP method is utilized to adjust 3D face model to be in the right front pose for a better recognition performance. By using the normalized facial depth image been acquired through the two previous steps, and by calculating the Gaussian and mean curvatures at each point, the surface types are classified and the classification result is used to mark different kinds of area on the facial depth image by 8 gray-levels. This achieved gray image is named as Surface Classification Image (SCI) and the SCI now represents the 3D features of the face and then it is input to the process of PCA to obtain the SCI eigenfaces to recognize the face. In the experiments conducted on 3D Facial database ZJU-3DFED of Zhejiang University, we obtained the rank-1 identification score of 94.5%, which outperformed the result of using PCA method directly on the face depth image (instead of SCI) by 16.5%.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125524035","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Automatic analysis of facial expression has become a popular research area because of it’s many applications in the field of computer vision. This paper presents a hybrid method based on Gabor filter, Kernel Principle Component Analysis (KPCA) and Support Vector Machine (SVM) for classification of facial expressions into six basic emotions. At first, Gabor filter bank is applied on input images. Then, the feature reduction technique of KPCA is performed on the outputs of the filter. Finally, SVM is used for classification. The proposed method is tested on the Cohen-Kanade’s facial expression images dataset. The results of the proposed method are compared to the ones of the combined Principle Component Analysis (PCA) and SVM classifier. Experimental results show the effectiveness of the proposed method. The average recognition rate of 89.9% is achieved in this work which is higher than 87.3% resulted from a common combined PCA and SVM method.
{"title":"A Combined KPCA and SVM Method for Basic Emotional Expressions Recognition","authors":"S. Fazli, R. Afrouzian, Hadi Seyedarabi","doi":"10.1109/ICMV.2009.67","DOIUrl":"https://doi.org/10.1109/ICMV.2009.67","url":null,"abstract":"Automatic analysis of facial expression has become a popular research area because of it’s many applications in the field of computer vision. This paper presents a hybrid method based on Gabor filter, Kernel Principle Component Analysis (KPCA) and Support Vector Machine (SVM) for classification of facial expressions into six basic emotions. At first, Gabor filter bank is applied on input images. Then, the feature reduction technique of KPCA is performed on the outputs of the filter. Finally, SVM is used for classification. The proposed method is tested on the Cohen-Kanade’s facial expression images dataset. The results of the proposed method are compared to the ones of the combined Principle Component Analysis (PCA) and SVM classifier. Experimental results show the effectiveness of the proposed method. The average recognition rate of 89.9% is achieved in this work which is higher than 87.3% resulted from a common combined PCA and SVM method.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125525259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Contourlet transform is a new multiscale and multidirectional image representation which effectively captures the edges and contours of images. Hidden Markov Tree model (HMT) can capture all inter-scale, interdirection and inter-location dependencies. Also, HMT can capture the statistical properties of the contourlet coefficients. Therefore, it is used to detect the image singularities (edges and ridges). In this paper, we have proposed three methods for texture segmentation, based on the HMT contourlet model. At first contourlet coefficient is computed and then, for each texture an HMT Contourlet model is trained for test phase, a set of decisions are made for each block of input image based on the maximum likelihood probability. Final decision will be based on the majority vote criterion. The proposed method has been examined on test images and promising results in terms of low segmentation errors has been obtained.
{"title":"Hmt-Contourlet Image Segmentation Based on Majority Vote","authors":"M. Helfroush, Narges Taghdir","doi":"10.1109/ICMV.2009.60","DOIUrl":"https://doi.org/10.1109/ICMV.2009.60","url":null,"abstract":"Contourlet transform is a new multiscale and multidirectional image representation which effectively captures the edges and contours of images. Hidden Markov Tree model (HMT) can capture all inter-scale, interdirection and inter-location dependencies. Also, HMT can capture the statistical properties of the contourlet coefficients. Therefore, it is used to detect the image singularities (edges and ridges). In this paper, we have proposed three methods for texture segmentation, based on the HMT contourlet model. At first contourlet coefficient is computed and then, for each texture an HMT Contourlet model is trained for test phase, a set of decisions are made for each block of input image based on the maximum likelihood probability. Final decision will be based on the majority vote criterion. The proposed method has been examined on test images and promising results in terms of low segmentation errors has been obtained.","PeriodicalId":315778,"journal":{"name":"2009 Second International Conference on Machine Vision","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126573675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：