Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6744054
Di Ming, Q. Wen, Juan Chen, Wenhao Liu
Malignant melanoma is among the most rapidly increasing cancers in the world. Image border detection is often the first step to characterize skin lesion for the follow-up computer-aided diagnosis. Existing approaches lack robustness in the face of dermoscopy images varying in size, color, texture, and structure. In this paper, a generalized Markov random field (MRF) framework is proposed to fuse the results obtained from segmentation algorithms, by taking full advantages of characteristics of different methods and making them work synergistically to acquire more reliable results. The experimental results on the real dermoscopy image set demonstrate that the proposed fusion method is capable of improving the overall performance in terms of both accuracy and robustness.
{"title":"A generalized fusion approach for segmenting dermoscopy images using Markov random field","authors":"Di Ming, Q. Wen, Juan Chen, Wenhao Liu","doi":"10.1109/CISP.2013.6744054","DOIUrl":"https://doi.org/10.1109/CISP.2013.6744054","url":null,"abstract":"Malignant melanoma is among the most rapidly increasing cancers in the world. Image border detection is often the first step to characterize skin lesion for the follow-up computer-aided diagnosis. Existing approaches lack robustness in the face of dermoscopy images varying in size, color, texture, and structure. In this paper, a generalized Markov random field (MRF) framework is proposed to fuse the results obtained from segmentation algorithms, by taking full advantages of characteristics of different methods and making them work synergistically to acquire more reliable results. The experimental results on the real dermoscopy image set demonstrate that the proposed fusion method is capable of improving the overall performance in terms of both accuracy and robustness.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125563215","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6743952
Miao Xie, Jiankun Hu
Host-based intrusion detection systems (HIDSs), especially anomaly-based, have received much attention over the past few decades. Over time, however, the existing data sets used for evaluation of a HIDS have lost most of their relevance due to the substantial development of computer systems. To fill this gap, ADFA Linux data set (ADFA-LD) is recently released, which is composed of thousands of system call traces collected from a contemporary Linux local server and expects to be a new benchmark for evaluating a HIDS. In this paper, we perform a preliminary analysis of ADFA-LD, in an attempt to extract useful information for developing new host-based anomaly detection systems (HADSs). In accordance with the general concerns arising from the community, some typical features are analysed particularly against ADFA-LD, such as length, common pattern and frequency. Furthermore, we implement a simple k nearest neighbour (kNN)-based HADS to be evaluated using ADFA-LD. The experimental results show that, although an acceptable performance can be acquired for a few types of attack, there is still a long way to fully understand the complex behaviour resulting from a modern computer system and, finally, realise more intelligent HADSs.
{"title":"Evaluating host-based anomaly detection systems: A preliminary analysis of ADFA-LD","authors":"Miao Xie, Jiankun Hu","doi":"10.1109/CISP.2013.6743952","DOIUrl":"https://doi.org/10.1109/CISP.2013.6743952","url":null,"abstract":"Host-based intrusion detection systems (HIDSs), especially anomaly-based, have received much attention over the past few decades. Over time, however, the existing data sets used for evaluation of a HIDS have lost most of their relevance due to the substantial development of computer systems. To fill this gap, ADFA Linux data set (ADFA-LD) is recently released, which is composed of thousands of system call traces collected from a contemporary Linux local server and expects to be a new benchmark for evaluating a HIDS. In this paper, we perform a preliminary analysis of ADFA-LD, in an attempt to extract useful information for developing new host-based anomaly detection systems (HADSs). In accordance with the general concerns arising from the community, some typical features are analysed particularly against ADFA-LD, such as length, common pattern and frequency. Furthermore, we implement a simple k nearest neighbour (kNN)-based HADS to be evaluated using ADFA-LD. The experimental results show that, although an acceptable performance can be acquired for a few types of attack, there is still a long way to fully understand the complex behaviour resulting from a modern computer system and, finally, realise more intelligent HADSs.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126684160","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6743866
Lili Cao, Zhi-Shui Zhang, Jun Zhang
Gas chromatography-mass spectrometry (GC-MS) is one of the most important and powerful tools to identify compounds in both chemical and biological samples. In this work, a novel compound identification method based on the dynamic multiple spectral similarity measures is proposed. The proposed method uses seven spectral similarity measures. To reduce the computational time, DFTR measure is used a filter layer in proposed method. 22457 mass spectra for 15793 unique compounds are used as query data and NIST05 main spectral library is used as reference library. The experimental results showed that the identification accuracy of the dynamic multiple similarity measures is increased 8.97% and 18.46% comparing with DFTR and Correlation measure, respectively.
{"title":"Dynamic multiple spectral similarity measures for compound identification","authors":"Lili Cao, Zhi-Shui Zhang, Jun Zhang","doi":"10.1109/CISP.2013.6743866","DOIUrl":"https://doi.org/10.1109/CISP.2013.6743866","url":null,"abstract":"Gas chromatography-mass spectrometry (GC-MS) is one of the most important and powerful tools to identify compounds in both chemical and biological samples. In this work, a novel compound identification method based on the dynamic multiple spectral similarity measures is proposed. The proposed method uses seven spectral similarity measures. To reduce the computational time, DFTR measure is used a filter layer in proposed method. 22457 mass spectra for 15793 unique compounds are used as query data and NIST05 main spectral library is used as reference library. The experimental results showed that the identification accuracy of the dynamic multiple similarity measures is increased 8.97% and 18.46% comparing with DFTR and Correlation measure, respectively.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126343766","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6745255
Xin Liu, Qian Zhang, Ru-peng Luan, Feng Yu
According to the characteristics of agriculture image, new image recognition scheme and image authentication scheme were designed, which based on perceptual hash algorithm. Some tomato leaf diseases pictures were used to achieve image perceptual hash feature extraction. And the experimental results show that the same diseases images have closer perceptual hash characteristics.
{"title":"Applications of perceptual hash algorithm in agriculture images","authors":"Xin Liu, Qian Zhang, Ru-peng Luan, Feng Yu","doi":"10.1109/CISP.2013.6745255","DOIUrl":"https://doi.org/10.1109/CISP.2013.6745255","url":null,"abstract":"According to the characteristics of agriculture image, new image recognition scheme and image authentication scheme were designed, which based on perceptual hash algorithm. Some tomato leaf diseases pictures were used to achieve image perceptual hash feature extraction. And the experimental results show that the same diseases images have closer perceptual hash characteristics.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126432764","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6744034
N. Lu, Hongyu Miao
To analyze the characteristics of the blood vessels quantitatively is of great importance to the diagnosis of different diseases, e.g., stenosis, diabetic retinopathy, tumor and so on. However, the occurrence of imaging noise or illumination variation may include difficulty for image analysis, especially for accurate quantitative analysis. Therefore, more efficient algorithms for vessel image thresholding, segmentation or extraction have been studied. Nevertheless, these methods may still fail entirely in case of images with large noise or structural noise (such as spots that obscure and/or are brighter than vessels). To improve the algorithm performance on these issues, we propose a novel thresholding algorithm for vascular images by probing the polarity in the circular profiles (PCP) of image pixels. This can robustly distinguish tube-like objects from both cloud-like contaminations and structural noise. Extensive simulation studies based on multiple evaluation criteria suggest that the PCP algorithm typically has a superior performance over other representative approaches. Finally, we also demonstrate the satisfactory performance of the PCP method on real image data.
{"title":"Featured circular profile for vessel thresholding","authors":"N. Lu, Hongyu Miao","doi":"10.1109/CISP.2013.6744034","DOIUrl":"https://doi.org/10.1109/CISP.2013.6744034","url":null,"abstract":"To analyze the characteristics of the blood vessels quantitatively is of great importance to the diagnosis of different diseases, e.g., stenosis, diabetic retinopathy, tumor and so on. However, the occurrence of imaging noise or illumination variation may include difficulty for image analysis, especially for accurate quantitative analysis. Therefore, more efficient algorithms for vessel image thresholding, segmentation or extraction have been studied. Nevertheless, these methods may still fail entirely in case of images with large noise or structural noise (such as spots that obscure and/or are brighter than vessels). To improve the algorithm performance on these issues, we propose a novel thresholding algorithm for vascular images by probing the polarity in the circular profiles (PCP) of image pixels. This can robustly distinguish tube-like objects from both cloud-like contaminations and structural noise. Extensive simulation studies based on multiple evaluation criteria suggest that the PCP algorithm typically has a superior performance over other representative approaches. Finally, we also demonstrate the satisfactory performance of the PCP method on real image data.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121304059","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}
Recently some types of quasi-circular patchy vegetations have been found and proved by the high spatial resolution satellite remote sensing images and field investigations in the Yellow River Delta. In order to promote the monitoring of vegetation community patch extension and quantify the mechanism of vegetation community patch succession, there is a clear need for accurately and economically detecting the vegetation community patches. The main objective of this paper is to detect the quasi-circular vegetation community patches from the different spatial resolution images of the first civilian high resolution optical stereo mapping satellite of China (ZY-3). The results show that the front-facing, rear-facing, ground-facing panchromatic and multispectral images are effective and sufficient to detect the quasi-circular vegetation patches based on Circular Hough Transform (CHT), and the higher spatial resolution image had potential to provide the higher detection accuracy for the quasi-circular vegetation patches.
{"title":"Detect quasi-circular vegetation community patches using images of different spatial resolutions","authors":"Yuanyuan Li, Qingsheng Liu, Gaohuan Liu, Chong Huang","doi":"10.1109/CISP.2013.6745279","DOIUrl":"https://doi.org/10.1109/CISP.2013.6745279","url":null,"abstract":"Recently some types of quasi-circular patchy vegetations have been found and proved by the high spatial resolution satellite remote sensing images and field investigations in the Yellow River Delta. In order to promote the monitoring of vegetation community patch extension and quantify the mechanism of vegetation community patch succession, there is a clear need for accurately and economically detecting the vegetation community patches. The main objective of this paper is to detect the quasi-circular vegetation community patches from the different spatial resolution images of the first civilian high resolution optical stereo mapping satellite of China (ZY-3). The results show that the front-facing, rear-facing, ground-facing panchromatic and multispectral images are effective and sufficient to detect the quasi-circular vegetation patches based on Circular Hough Transform (CHT), and the higher spatial resolution image had potential to provide the higher detection accuracy for the quasi-circular vegetation patches.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122560665","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6743863
Hao Chen, Jun-hai Guo, Tong Liu
Combining EMD (Empirical mode decomposition) based denoising methods with mode decomposition methods, new robust FM (frequency modulation) signal parameter estimation methods are proposed. The original signal is firstly denoised with EMD based denoising method, HHT (Hilbert Huang transform) based FM signal parameter estimation method is then used on the denoised signal. Simulations show that the EMD based denoising method largely increases signal's SNR and does no harm to the mode decomposition, making the FM parameter estimation method more robust to noise, and the application in radar target estimation proves the usefulness of this method.
{"title":"FM signal parameter estimation using the Hilbert Huang transform","authors":"Hao Chen, Jun-hai Guo, Tong Liu","doi":"10.1109/CISP.2013.6743863","DOIUrl":"https://doi.org/10.1109/CISP.2013.6743863","url":null,"abstract":"Combining EMD (Empirical mode decomposition) based denoising methods with mode decomposition methods, new robust FM (frequency modulation) signal parameter estimation methods are proposed. The original signal is firstly denoised with EMD based denoising method, HHT (Hilbert Huang transform) based FM signal parameter estimation method is then used on the denoised signal. Simulations show that the EMD based denoising method largely increases signal's SNR and does no harm to the mode decomposition, making the FM parameter estimation method more robust to noise, and the application in radar target estimation proves the usefulness of this method.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122189201","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}
To segment piecewise smooth images on implicit surfaces, Potts model which is a general variational model for planar multiphase image segmentation is extended to the one for image segmentation on implicit surfaces. Potts model is reformulated not only on implicit closed surfaces but also on implicit open surfaces. An improved dual method is proposed to speed up the computation process. Some numerical experiments are presented to give qualitative and quantitative comparisons among our method and the other two fast methods, that is, Split Bregman method and dual method. The results show that these methods can all segment piecewise smooth images on implicit surfaces effectively while the improved dual method is more efficient than Split Bregman method and dual method.
{"title":"A multiphase image segmentation model for piecewise smooth images on implicit surfaces","authors":"Jing Wang, Yong-guo Zheng, Zhenkuan Pan, Wei Zhong Zhang, J. Duan","doi":"10.1109/CISP.2013.6744055","DOIUrl":"https://doi.org/10.1109/CISP.2013.6744055","url":null,"abstract":"To segment piecewise smooth images on implicit surfaces, Potts model which is a general variational model for planar multiphase image segmentation is extended to the one for image segmentation on implicit surfaces. Potts model is reformulated not only on implicit closed surfaces but also on implicit open surfaces. An improved dual method is proposed to speed up the computation process. Some numerical experiments are presented to give qualitative and quantitative comparisons among our method and the other two fast methods, that is, Split Bregman method and dual method. The results show that these methods can all segment piecewise smooth images on implicit surfaces effectively while the improved dual method is more efficient than Split Bregman method and dual method.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115315102","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6745252
Lipin Lu, Yuanyuan Pu, Heng Zhang, Dan Xu
In this paper, a non-photorealistic rendering technology of cartoons is studied. A new algorithm which can automatically generate cartoon paintings is proposed. This method needs no user's participation and no drawing skills. Firstly, by using the bilateral filtering, the abstract visually appealing and restrain minor features can be highlighted. Second, the abstract image is quantified to imitate the coloring style of the cartoon. Third, FDOG filter is used to draw contour lines of the abstract image to further improve visualization. Last, the quantized image and the contour image are fused to produce an image with the cartoons. The result shows that cartoons generated by this algorithm have characters of art aesthetic and a better visualization.
{"title":"A non-photorealistic rendering algorithm for cartoons","authors":"Lipin Lu, Yuanyuan Pu, Heng Zhang, Dan Xu","doi":"10.1109/CISP.2013.6745252","DOIUrl":"https://doi.org/10.1109/CISP.2013.6745252","url":null,"abstract":"In this paper, a non-photorealistic rendering technology of cartoons is studied. A new algorithm which can automatically generate cartoon paintings is proposed. This method needs no user's participation and no drawing skills. Firstly, by using the bilateral filtering, the abstract visually appealing and restrain minor features can be highlighted. Second, the abstract image is quantified to imitate the coloring style of the cartoon. Third, FDOG filter is used to draw contour lines of the abstract image to further improve visualization. Last, the quantized image and the contour image are fused to produce an image with the cartoons. The result shows that cartoons generated by this algorithm have characters of art aesthetic and a better visualization.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124970354","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}
Pub Date : 2013-12-01DOI: 10.1109/CISP.2013.6743907
Lei Kong, Zhijun Xu, Jinming Wang, Kexiu Pan
The jamming is one of the most important factors in modern wireless communication. More and more attention has been paid to the analysis of jamming signals. In this paper, a novel algorithm for the recognition of jamming signals in wireless communication is proposed. This algorithm is based on the singular value decomposition of the signals' cyclic spectrum density. The cyclic spectrum density is insensitive to Gauss white noise and put up many important signal features. The singular value decomposition method is conducive to decrease the dimensions of cyclic spectrum density function. Simulation results show that the proposed algorithm has a higher recognition rate than the traditional method, especially when the jamming to signal ratio is lower.
{"title":"A novel algorithm for jamming recognition in wireless communication","authors":"Lei Kong, Zhijun Xu, Jinming Wang, Kexiu Pan","doi":"10.1109/CISP.2013.6743907","DOIUrl":"https://doi.org/10.1109/CISP.2013.6743907","url":null,"abstract":"The jamming is one of the most important factors in modern wireless communication. More and more attention has been paid to the analysis of jamming signals. In this paper, a novel algorithm for the recognition of jamming signals in wireless communication is proposed. This algorithm is based on the singular value decomposition of the signals' cyclic spectrum density. The cyclic spectrum density is insensitive to Gauss white noise and put up many important signal features. The singular value decomposition method is conducive to decrease the dimensions of cyclic spectrum density function. Simulation results show that the proposed algorithm has a higher recognition rate than the traditional method, especially when the jamming to signal ratio is lower.","PeriodicalId":442320,"journal":{"name":"2013 6th International Congress on Image and Signal Processing (CISP)","volume":"131 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132369128","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：