Pub Date : 2013-10-01DOI: 10.1109/IST.2013.6729726
Ruofan Wang, Liwen Miao, Yixin Ma
With the development of electronic technology and image reconstruction algorithm, the quality of EIT image is improved significantly. EIT technology has the potential to be of great value in medical and industrial applications. However, there is very little research about the estimation of anomaly's boundary from EIT image. The gradient method is proposed in this paper to identify object boundary from EIT image reconstructed by sensitivity conjugate gradient (SCG) algorithm. The performance of gradient method in terms of position error and size error is compared with traditional threshold method at two threshold levels through computer simulation and phantom experiments. The results show that the result of gradient method can cover the anomaly region in investigated situations and rely on very little priori-knowledge, while threshold method can estimate more accurately if the threshold value is set properly based on priori-knowledge. Improper threshold definition may lead to significant error in boundary identification. The pilot study to estimate the location and size of the anomaly from EIT image further completes EIT research and promotes its practical application.
{"title":"A gradient method used to identify object boundary in EIT image","authors":"Ruofan Wang, Liwen Miao, Yixin Ma","doi":"10.1109/IST.2013.6729726","DOIUrl":"https://doi.org/10.1109/IST.2013.6729726","url":null,"abstract":"With the development of electronic technology and image reconstruction algorithm, the quality of EIT image is improved significantly. EIT technology has the potential to be of great value in medical and industrial applications. However, there is very little research about the estimation of anomaly's boundary from EIT image. The gradient method is proposed in this paper to identify object boundary from EIT image reconstructed by sensitivity conjugate gradient (SCG) algorithm. The performance of gradient method in terms of position error and size error is compared with traditional threshold method at two threshold levels through computer simulation and phantom experiments. The results show that the result of gradient method can cover the anomaly region in investigated situations and rely on very little priori-knowledge, while threshold method can estimate more accurately if the threshold value is set properly based on priori-knowledge. Improper threshold definition may lead to significant error in boundary identification. The pilot study to estimate the location and size of the anomaly from EIT image further completes EIT research and promotes its practical application.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125823740","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 : 2013-10-01DOI: 10.1109/IST.2013.6729698
Muhammad Zubair, Abubakar Yamin, S. Khan
Optic Disc (OD) behaves as a landmark in a fundus image for the localization of other major parts of the eye like macula region, fovea and the network of blood vessels. OD can be seen as a quite distinct bright region in a colored fundus image. In our proposed method the high intensity value of the optic disc is used for its detection. First the optic disc is detected by applying preprocessing steps for enhancement like Contrast-Limited Adaptive Histogram Equalization (CLAHE), contrast stretching transformation and extended minima transformation. OD free image is obtained by applying morphological opening and closing operations on the preprocessed image. A sensitivity, specificity and accuracy of 100%, 98.25% and 98.65% is achieved respectively on 1200 images from publically available database MESSIDOR. Compared with state of the art, the proposed technique has reasonable accuracy using less computing power and processing time.
{"title":"Automated detection of Optic Disc for the analysis of retina using color fundus image","authors":"Muhammad Zubair, Abubakar Yamin, S. Khan","doi":"10.1109/IST.2013.6729698","DOIUrl":"https://doi.org/10.1109/IST.2013.6729698","url":null,"abstract":"Optic Disc (OD) behaves as a landmark in a fundus image for the localization of other major parts of the eye like macula region, fovea and the network of blood vessels. OD can be seen as a quite distinct bright region in a colored fundus image. In our proposed method the high intensity value of the optic disc is used for its detection. First the optic disc is detected by applying preprocessing steps for enhancement like Contrast-Limited Adaptive Histogram Equalization (CLAHE), contrast stretching transformation and extended minima transformation. OD free image is obtained by applying morphological opening and closing operations on the preprocessed image. A sensitivity, specificity and accuracy of 100%, 98.25% and 98.65% is achieved respectively on 1200 images from publically available database MESSIDOR. Compared with state of the art, the proposed technique has reasonable accuracy using less computing power and processing time.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116543283","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 : 2013-10-01DOI: 10.1109/IST.2013.6729724
W. Smolik, J. Kryszyn
This paper presents modeling of electric potential distribution using a refined Cartesian mesh in electrical capacitance tomography. The application of the thickened Cartesian discretization mesh in a forward problem solver allows more accurate discrete description of cylindrical elements of tomographic sensor. The irregular structured grid enables possibility to increase precision of capacitance measurement simulations. A tree structure was used for description of refined mesh nodes to obtain fast computations. The .NET based software library was created. A set of MATLAB functions was created to provide an interface between the library and the MATLAB environment. The results of electrical field computation were presented.
{"title":"Refined cartesian mesh for modeling in electrical capacitance tomography","authors":"W. Smolik, J. Kryszyn","doi":"10.1109/IST.2013.6729724","DOIUrl":"https://doi.org/10.1109/IST.2013.6729724","url":null,"abstract":"This paper presents modeling of electric potential distribution using a refined Cartesian mesh in electrical capacitance tomography. The application of the thickened Cartesian discretization mesh in a forward problem solver allows more accurate discrete description of cylindrical elements of tomographic sensor. The irregular structured grid enables possibility to increase precision of capacitance measurement simulations. A tree structure was used for description of refined mesh nodes to obtain fast computations. The .NET based software library was created. A set of MATLAB functions was created to provide an interface between the library and the MATLAB environment. The results of electrical field computation were presented.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"96 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115607857","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 : 2013-10-01DOI: 10.1109/IST.2013.6729716
Zhaoguo Wu, Ya Zhou, Xiaoming Hu, Muqing Zhou, Xiaobing Dai, Xinzhou Li, Danting Wang
Subcutaneous vein is invisible to naked eyes, but can be easily identified under NIR (near-infrared) light, which is widely used in the catheter insertion and biometric identification. However, the quality of raw NIR image usually suffers from low contrast due to uneven illumination and individual difference. Most of current methods used to enhance the contrast between veins and surrounding tissue are based on single wavelength NIR LED. In the meaning the difference of individual NIR absorption is ignored, which results in diverse performance among different individuals. In this work, a training-based contrast enhancement algorithm is applied to hand vein images. NiBlack segmentation method is also used in the NIR system in order to get a high contrast binary image. An enhanced NIR image is generated from processing six images acquired under six mono-wavelength of light (730nm, 830nm, 850nm, 880nm, 890nm and 940nm). The experiment shows that the contrast of the enhanced NIR image increased by 4~10 times.
{"title":"A vein image enhancement algorithm for the multi-spectral illumination","authors":"Zhaoguo Wu, Ya Zhou, Xiaoming Hu, Muqing Zhou, Xiaobing Dai, Xinzhou Li, Danting Wang","doi":"10.1109/IST.2013.6729716","DOIUrl":"https://doi.org/10.1109/IST.2013.6729716","url":null,"abstract":"Subcutaneous vein is invisible to naked eyes, but can be easily identified under NIR (near-infrared) light, which is widely used in the catheter insertion and biometric identification. However, the quality of raw NIR image usually suffers from low contrast due to uneven illumination and individual difference. Most of current methods used to enhance the contrast between veins and surrounding tissue are based on single wavelength NIR LED. In the meaning the difference of individual NIR absorption is ignored, which results in diverse performance among different individuals. In this work, a training-based contrast enhancement algorithm is applied to hand vein images. NiBlack segmentation method is also used in the NIR system in order to get a high contrast binary image. An enhanced NIR image is generated from processing six images acquired under six mono-wavelength of light (730nm, 830nm, 850nm, 880nm, 890nm and 940nm). The experiment shows that the contrast of the enhanced NIR image increased by 4~10 times.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"22 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126058707","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 : 2013-10-01DOI: 10.1109/IST.2013.6729700
T. C. Silva, Liang Zhao
Complex networks refer to large-scale graphs with nontrivial connection patterns. The salient and interesting features that the complex network study offers in comparison to graph theory are the emphasis on the dynamical properties of the networks and the ability of inherently uncovering pattern formation of the vertices. In this paper, we present a hybrid data classification technique combining a low level and a high level classifier. The low level term can be equipped with any traditional classification techniques, which realize the classification task considering only physical or topological features (e.g., geometrical or statistical features) of the input data. On the other hand, the high level term has the ability of detecting data patterns with semantic meanings. In this way, the classification is realized by means of the extraction of the underlying network's features constructed from the input data. As a result, the high level classification process measures the compliance of the test instances with the pattern formation of the training data. Out of various high level perspectives that can be utilized to capture semantic meaning, we utilize the dynamical features that are generated from a tourist walker in a networked environment. Specifically, a weighted combination of transient and cycle lengths generated by the tourist walk is employed for that end. A study on the critical memory length is provided. Finally, we apply the proposed technique to the recognition of handwritten digit images and promising results have been obtained.
{"title":"Handwritten digits recognition using a high level network-based approach","authors":"T. C. Silva, Liang Zhao","doi":"10.1109/IST.2013.6729700","DOIUrl":"https://doi.org/10.1109/IST.2013.6729700","url":null,"abstract":"Complex networks refer to large-scale graphs with nontrivial connection patterns. The salient and interesting features that the complex network study offers in comparison to graph theory are the emphasis on the dynamical properties of the networks and the ability of inherently uncovering pattern formation of the vertices. In this paper, we present a hybrid data classification technique combining a low level and a high level classifier. The low level term can be equipped with any traditional classification techniques, which realize the classification task considering only physical or topological features (e.g., geometrical or statistical features) of the input data. On the other hand, the high level term has the ability of detecting data patterns with semantic meanings. In this way, the classification is realized by means of the extraction of the underlying network's features constructed from the input data. As a result, the high level classification process measures the compliance of the test instances with the pattern formation of the training data. Out of various high level perspectives that can be utilized to capture semantic meaning, we utilize the dynamical features that are generated from a tourist walker in a networked environment. Specifically, a weighted combination of transient and cycle lengths generated by the tourist walk is employed for that end. A study on the critical memory length is provided. Finally, we apply the proposed technique to the recognition of handwritten digit images and promising results have been obtained.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"133 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121497787","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 : 2013-10-01DOI: 10.1109/IST.2013.6729662
Yong Zhang, Sanxia Xie, Shulin Wei
A novel descriptor for laser printing character recognition is proposed by using the fusion features and multilevel classification. There are two level features in the feature extraction module, the first level feature uses the coarse grid statistical features, having great applicability for local character distortion and stroke thickness inequality. Wavelet moment, as the second level feature, has the scale, translation and rotation invariance and the great anti-noise capability. It can reflect the approximation of the overall character and local detail information at the same time, so it is suitable for classing similar characters. As for the classification module, based on results of primary feature template matching classification in the first place, the second matching uses two level features fusion and special distortion character template to classify the top three character in the results. Experimental results demonstrate that using the fusion features and multilevel classification, industrial laser character recognition rate can be up to 99.2%, which is better than that of using single stage feature recognition.
{"title":"Industrial character recognition based on grid feature and wavelet moment","authors":"Yong Zhang, Sanxia Xie, Shulin Wei","doi":"10.1109/IST.2013.6729662","DOIUrl":"https://doi.org/10.1109/IST.2013.6729662","url":null,"abstract":"A novel descriptor for laser printing character recognition is proposed by using the fusion features and multilevel classification. There are two level features in the feature extraction module, the first level feature uses the coarse grid statistical features, having great applicability for local character distortion and stroke thickness inequality. Wavelet moment, as the second level feature, has the scale, translation and rotation invariance and the great anti-noise capability. It can reflect the approximation of the overall character and local detail information at the same time, so it is suitable for classing similar characters. As for the classification module, based on results of primary feature template matching classification in the first place, the second matching uses two level features fusion and special distortion character template to classify the top three character in the results. Experimental results demonstrate that using the fusion features and multilevel classification, industrial laser character recognition rate can be up to 99.2%, which is better than that of using single stage feature recognition.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128711577","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 : 2013-10-01DOI: 10.1109/IST.2013.6729729
M. Bae, S. Jo, Hee-Ho Lee, Minho Lee, Juyeong Kim, P. Choi, Jang-Kyoo Shin, Seung Soo Kim, Jeongyeob Kim
In this study, we propose a new complementary metal-oxide semiconductor (CMOS) image sensor to extend the dynamic range. The proposed active pixel sensor (APS) uses a gate/body-tied PMOSFET-type (GBT) photodetector and an output voltage feedback structure. Although the size of the proposed APS is slightly larger than a conventional APS, the dynamic range of the proposed pixel is much wider than a conventional APS. And the proposed APS has a varied dynamic range from adjusting the reference voltage. The proposed APS has been designed and fabricated by using 0.35 μm 2-poly 4-metal standard CMOS technology, and its characteristics have been evaluated.
{"title":"A novel CMOS image sensor using an output voltage feedback with variable dynamic range","authors":"M. Bae, S. Jo, Hee-Ho Lee, Minho Lee, Juyeong Kim, P. Choi, Jang-Kyoo Shin, Seung Soo Kim, Jeongyeob Kim","doi":"10.1109/IST.2013.6729729","DOIUrl":"https://doi.org/10.1109/IST.2013.6729729","url":null,"abstract":"In this study, we propose a new complementary metal-oxide semiconductor (CMOS) image sensor to extend the dynamic range. The proposed active pixel sensor (APS) uses a gate/body-tied PMOSFET-type (GBT) photodetector and an output voltage feedback structure. Although the size of the proposed APS is slightly larger than a conventional APS, the dynamic range of the proposed pixel is much wider than a conventional APS. And the proposed APS has a varied dynamic range from adjusting the reference voltage. The proposed APS has been designed and fabricated by using 0.35 μm 2-poly 4-metal standard CMOS technology, and its characteristics have been evaluated.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124564244","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 : 2013-10-01DOI: 10.1109/IST.2013.6729655
Jiamin Ye, Haigang Wang, Wuqiang Yang
It is assumed that the spatial resolution of the reconstructed images with an electrical resistance tomography (ERT) system can be improved by increasing the number of measurement electrodes. In this paper, the issues with the number of electrodes will be discussed. To analyze the effect of the number of electrodes on the quality of reconstructed images, ERT sensors with different number of electrodes-8, 12, 16, 20, 24, 32, 36, 40 and 48-are investigated. With the adjacent and opposite strategies, the voltages between different electrode pairs are calculated for some typical conductivity distributions with different contrast in conductivity using a finite element method. The obtained voltage data are then used to reconstruct images using the projected Landweber iteration algorithm. In the meanwhile, the impact of meshes in the inverse problem solver on image quality is also discussed. The sensitivity distributions for all cases are analyzed utilizing singular value decomposition (SVD). The analytical results are in agreement with the simulation results.
{"title":"Evaluation of effect of number of electrodes in ERT sensors on image quality","authors":"Jiamin Ye, Haigang Wang, Wuqiang Yang","doi":"10.1109/IST.2013.6729655","DOIUrl":"https://doi.org/10.1109/IST.2013.6729655","url":null,"abstract":"It is assumed that the spatial resolution of the reconstructed images with an electrical resistance tomography (ERT) system can be improved by increasing the number of measurement electrodes. In this paper, the issues with the number of electrodes will be discussed. To analyze the effect of the number of electrodes on the quality of reconstructed images, ERT sensors with different number of electrodes-8, 12, 16, 20, 24, 32, 36, 40 and 48-are investigated. With the adjacent and opposite strategies, the voltages between different electrode pairs are calculated for some typical conductivity distributions with different contrast in conductivity using a finite element method. The obtained voltage data are then used to reconstruct images using the projected Landweber iteration algorithm. In the meanwhile, the impact of meshes in the inverse problem solver on image quality is also discussed. The sensitivity distributions for all cases are analyzed utilizing singular value decomposition (SVD). The analytical results are in agreement with the simulation results.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114470393","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 : 2013-10-01DOI: 10.1109/IST.2013.6729685
A. Skalski, P. Kedzierawski, T. Zielinski, T. Kuszewski
This paper addresses a problem of automatic segmentation of computed tomography (CT) data in context of prostate radiotherapy planning. A new 3D algorithm is proposed in which a prostate is automatically contoured in CT images. The proposed segmentation scenario consists of the following steps: 1) both CT and magnetic resonance (MR) data of a patient are acquired, 2) due to better visibility of soft tissues in MR images, soft organs are segmented in MR data using active contour method (snakes) with additional gradient vector flow enhancement, 3) then obtained 3D contours are mapped from MR to CT images (using mutual information criterion) by means of a flexible registration technique in which global affine transformation is combined with local B-spline free from deformation method. During segmentation of the MR images prior knowledge about a mean ellipsoidal prostate shape, extracted before, plays a role of an addition constraint. Obtained results are compared with manual segmentation done by medical doctors using Dice similarity measure.
{"title":"CT segmentation based on MRI images in context of prostate radiotherapy planning","authors":"A. Skalski, P. Kedzierawski, T. Zielinski, T. Kuszewski","doi":"10.1109/IST.2013.6729685","DOIUrl":"https://doi.org/10.1109/IST.2013.6729685","url":null,"abstract":"This paper addresses a problem of automatic segmentation of computed tomography (CT) data in context of prostate radiotherapy planning. A new 3D algorithm is proposed in which a prostate is automatically contoured in CT images. The proposed segmentation scenario consists of the following steps: 1) both CT and magnetic resonance (MR) data of a patient are acquired, 2) due to better visibility of soft tissues in MR images, soft organs are segmented in MR data using active contour method (snakes) with additional gradient vector flow enhancement, 3) then obtained 3D contours are mapped from MR to CT images (using mutual information criterion) by means of a flexible registration technique in which global affine transformation is combined with local B-spline free from deformation method. During segmentation of the MR images prior knowledge about a mean ellipsoidal prostate shape, extracted before, plays a role of an addition constraint. Obtained results are compared with manual segmentation done by medical doctors using Dice similarity measure.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126804167","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 : 2013-10-01DOI: 10.1109/IST.2013.6729712
J. Ke, Ping Wei, Xin Zhang, E. Lam
In this paper, block-based compressive low-light-level imaging (BCL-imaging) is studied. To obtain larger measurement SNR (signal to noise ratio), instead of object pixels, linear combinations of pixels, referred to as features, are collected. PCA and Hadamard features are studied. Measurement SNR and reconstruction error are analyzed to quantify BCL-imaging performance. Compared with conventional imaging, BCL-imaging presents better reconstruction quality. Between PCA and Hadamard projections, PCA has smaller reconstruction error. However, after sorting the projection vectors using measurement SNR, Hadamard can obtain similarly performance as PCA. Biased vector and dual-measurements are studied with experimental results for the implementation of both projections in the end of this paper.
{"title":"Block-based compressive low-light-level imaging","authors":"J. Ke, Ping Wei, Xin Zhang, E. Lam","doi":"10.1109/IST.2013.6729712","DOIUrl":"https://doi.org/10.1109/IST.2013.6729712","url":null,"abstract":"In this paper, block-based compressive low-light-level imaging (BCL-imaging) is studied. To obtain larger measurement SNR (signal to noise ratio), instead of object pixels, linear combinations of pixels, referred to as features, are collected. PCA and Hadamard features are studied. Measurement SNR and reconstruction error are analyzed to quantify BCL-imaging performance. Compared with conventional imaging, BCL-imaging presents better reconstruction quality. Between PCA and Hadamard projections, PCA has smaller reconstruction error. However, after sorting the projection vectors using measurement SNR, Hadamard can obtain similarly performance as PCA. Biased vector and dual-measurements are studied with experimental results for the implementation of both projections in the end of this paper.","PeriodicalId":448698,"journal":{"name":"2013 IEEE International Conference on Imaging Systems and Techniques (IST)","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122556823","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：