Many state-of-the-art object recognition systems rely on identifying the location of objects in images, in order to better learn its visual attributes. In this paper, we propose four simple yet powerful hybrid ROI detection methods (combining both local and global features), based on frequently occurring keypoints. We show that our methods demonstrate competitive performance in two different types of datasets, the Caltech101 dataset and the GRAZ-02 dataset, where the pairs of keypoint bounding box method achieved the best accuracies overall.
{"title":"SIFTing the Relevant from the Irrelevant: Automatically Detecting Objects in Training Images","authors":"E. Zhang, M. Mayo","doi":"10.1109/DICTA.2009.59","DOIUrl":"https://doi.org/10.1109/DICTA.2009.59","url":null,"abstract":"Many state-of-the-art object recognition systems rely on identifying the location of objects in images, in order to better learn its visual attributes. In this paper, we propose four simple yet powerful hybrid ROI detection methods (combining both local and global features), based on frequently occurring keypoints. We show that our methods demonstrate competitive performance in two different types of datasets, the Caltech101 dataset and the GRAZ-02 dataset, where the pairs of keypoint bounding box method achieved the best accuracies overall.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123898450","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}
Yuee Liu, Zhengrong Li, R. Hayward, R. Walker, Hang Jin
Light Detection and Ranging (LIDAR) has great potential to assist vegetation management in power line corridors by providing more accurate geometric information of the power line assets and vegetation along the corridors. However, the development of algorithms for the automatic processing of LIDAR point cloud data, in particular for feature extraction and classification of raw point cloud data, is in still in its infancy. In this paper, we take advantage of LIDAR intensity and try to classify ground and non-ground points by statistically analyzing the skewness and kurtosis of the intensity data. Moreover, the Hough transform is employed to detected power lines from the filtered object points. The experimental results show the effectiveness of our methods and indicate that better results were obtained by using LIDAR intensity data than elevation data.
{"title":"Classification of Airborne LIDAR Intensity Data Using Statistical Analysis and Hough Transform with Application to Power Line Corridors","authors":"Yuee Liu, Zhengrong Li, R. Hayward, R. Walker, Hang Jin","doi":"10.1109/DICTA.2009.83","DOIUrl":"https://doi.org/10.1109/DICTA.2009.83","url":null,"abstract":"Light Detection and Ranging (LIDAR) has great potential to assist vegetation management in power line corridors by providing more accurate geometric information of the power line assets and vegetation along the corridors. However, the development of algorithms for the automatic processing of LIDAR point cloud data, in particular for feature extraction and classification of raw point cloud data, is in still in its infancy. In this paper, we take advantage of LIDAR intensity and try to classify ground and non-ground points by statistically analyzing the skewness and kurtosis of the intensity data. Moreover, the Hough transform is employed to detected power lines from the filtered object points. The experimental results show the effectiveness of our methods and indicate that better results were obtained by using LIDAR intensity data than elevation data.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115427300","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}
We are interested in the problem of automatically tracking football players, subject to the constraint that only one vantage point is available. Tracking algorithms benefit from seeing the entire playing field, as one does not have to worry about objects entering and leaving the field of view. However, the image of the entire field must be of sufficient resolution to allow each of the players to be identified automatically. To achieve this desired video data, several high definition video cameras are used to record a football match from a single vantage point. The cameras are oriented to cover the entire playing field, and their images combined to create a single high-resolution video feed. The user is able to pan and zoom in real-time within the unified video stream while it is playing. The system is achieved by distributing tasks across a network of computers and only processing data that will be visible to the user.
{"title":"Portable Multi-megapixel Camera with Real-Time Recording and Playback","authors":"Peter Carr, R. Hartley","doi":"10.1109/DICTA.2009.62","DOIUrl":"https://doi.org/10.1109/DICTA.2009.62","url":null,"abstract":"We are interested in the problem of automatically tracking football players, subject to the constraint that only one vantage point is available. Tracking algorithms benefit from seeing the entire playing field, as one does not have to worry about objects entering and leaving the field of view. However, the image of the entire field must be of sufficient resolution to allow each of the players to be identified automatically. To achieve this desired video data, several high definition video cameras are used to record a football match from a single vantage point. The cameras are oriented to cover the entire playing field, and their images combined to create a single high-resolution video feed. The user is able to pan and zoom in real-time within the unified video stream while it is playing. The system is achieved by distributing tasks across a network of computers and only processing data that will be visible to the user.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116637626","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}
Weihong Wang, Chunhua Shen, Jian Zhang, S. Paisitkriangkrai
We present a two-layer night time vehicle detector in this work. At the first layer, vehicle headlight detection is applied to find areas (bounding boxes) where the possible pairs of headlights locate in the image, the Haar feature based AdaBoost framework is then applied to detect the vehicle front. This approach has achieved a very promising performance for vehicle detection at night time. Our results show that the proposed algorithm can obtain a detection rate of over 90% at a very low false positive rate (1.5%). Without any code optimization, it also performs at a faster speed compared to the standard Haar feature based AdaBoost approach.
{"title":"A Two-Layer Night-Time Vehicle Detector","authors":"Weihong Wang, Chunhua Shen, Jian Zhang, S. Paisitkriangkrai","doi":"10.1109/DICTA.2009.33","DOIUrl":"https://doi.org/10.1109/DICTA.2009.33","url":null,"abstract":"We present a two-layer night time vehicle detector in this work. At the first layer, vehicle headlight detection is applied to find areas (bounding boxes) where the possible pairs of headlights locate in the image, the Haar feature based AdaBoost framework is then applied to detect the vehicle front. This approach has achieved a very promising performance for vehicle detection at night time. Our results show that the proposed algorithm can obtain a detection rate of over 90% at a very low false positive rate (1.5%). Without any code optimization, it also performs at a faster speed compared to the standard Haar feature based AdaBoost approach.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134496943","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}
A method based on sublevel sets is presented for refining segmentation of screening mammograms. Initial segmentation is provided by an adaptive pyramid (AP) scheme which is viewed as seeding of the final segmentation by sublevel sets. Performance is tested with and without prior anisotropic smoothing and is compared to refinement based on component merging. The combination of anisotropic smoothing, AP segmentation and sublevel refinement is found to outperform other combinations.
{"title":"Automatic Mass Segmentation Based on Adaptive Pyramid and Sublevel Set Analysis","authors":"Fei Ma, M. Bajger, M. Bottema","doi":"10.1109/DICTA.2009.47","DOIUrl":"https://doi.org/10.1109/DICTA.2009.47","url":null,"abstract":"A method based on sublevel sets is presented for refining segmentation of screening mammograms. Initial segmentation is provided by an adaptive pyramid (AP) scheme which is viewed as seeding of the final segmentation by sublevel sets. Performance is tested with and without prior anisotropic smoothing and is compared to refinement based on component merging. The combination of anisotropic smoothing, AP segmentation and sublevel refinement is found to outperform other combinations.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"99 6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134287427","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}
F. She, R. H. Chen, W. Gao, P. Hodgson, L. Kong, H. Hong
In this study, we focused on developing a novel 3D Thinning algorithm to extract one-voxel wide skeleton from various 3D objects aiming at preserving the topological information. The 3D Thinning algorithm was testified on computer-generated and real 3D reconstructed image sets acquired from TEMT and compared with other existing 3D Thinning algorithms. It is found that the algorithm has conserved medial axes and simultaneously topologies very well, demonstrating many advantages over the existing technologies. They are versatile, rigorous, efficient and rotation invariant.
{"title":"Improved 3D Thinning Algorithms for Skeleton Extraction","authors":"F. She, R. H. Chen, W. Gao, P. Hodgson, L. Kong, H. Hong","doi":"10.1109/DICTA.2009.13","DOIUrl":"https://doi.org/10.1109/DICTA.2009.13","url":null,"abstract":"In this study, we focused on developing a novel 3D Thinning algorithm to extract one-voxel wide skeleton from various 3D objects aiming at preserving the topological information. The 3D Thinning algorithm was testified on computer-generated and real 3D reconstructed image sets acquired from TEMT and compared with other existing 3D Thinning algorithms. It is found that the algorithm has conserved medial axes and simultaneously topologies very well, demonstrating many advantages over the existing technologies. They are versatile, rigorous, efficient and rotation invariant.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130583105","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}
M. Obaid, R. Mukundan, Hartmut Goecke, M. Billinghurst, H. Seichter
In this paper, we propose a novel approach for recognizing facial expressions based on using an Active Appearance Model facial feature tracking system with the quadratic deformation model representations of facial expressions. Thirty seven Facial Feature points are tracked based on the MPEG-4 Facial Animation Parameters layout. The proposed approach relies on the Euclidean distance measures between the tracked feature points and the reference deformed facial feature points of the six main expressions (smile, sad, fear, disgust, surprise, and anger). An evaluation of 30 model subjects, selected randomly from the Cohn-Kanade Database, was carried out. Results show that the main six facial expressions can successfully be recognized with an overall recognition accuracy of 89%. The proposed approach yields to promising recognition rates and can be used in real time applications.
{"title":"A Quadratic Deformation Model for Facial Expression Recognition","authors":"M. Obaid, R. Mukundan, Hartmut Goecke, M. Billinghurst, H. Seichter","doi":"10.1109/DICTA.2009.51","DOIUrl":"https://doi.org/10.1109/DICTA.2009.51","url":null,"abstract":"In this paper, we propose a novel approach for recognizing facial expressions based on using an Active Appearance Model facial feature tracking system with the quadratic deformation model representations of facial expressions. Thirty seven Facial Feature points are tracked based on the MPEG-4 Facial Animation Parameters layout. The proposed approach relies on the Euclidean distance measures between the tracked feature points and the reference deformed facial feature points of the six main expressions (smile, sad, fear, disgust, surprise, and anger). An evaluation of 30 model subjects, selected randomly from the Cohn-Kanade Database, was carried out. Results show that the main six facial expressions can successfully be recognized with an overall recognition accuracy of 89%. The proposed approach yields to promising recognition rates and can be used in real time applications.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128594840","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}
Len Hamey, R. Connally, Simon Wong Too Yen, Thomas S. Lawson, J. Piper, J. Iredell
Fluorescence microscopy is a powerful tool for the rapid identification of target organisms. However, natural autofluorescence often interferes with identification. Time-gated luminescence microscopy (TGLM) uses luminescent labels with long persistence in conjunction with digital imaging to regain discriminative power. Following the excitation pulse, short-lived autofluorescence decays rapidly whereas the long-lived emission from lanthanide doped polymer beads persists for hundreds of microseconds. After a short resolving period, a gated high gain camera captures the persistent emission in the absence of short-lived fluorescence. We report on the development of a TGLM software system for automated scanning of microscope slides, and show its use to resolve luminescent microspheres within a matrix of autofluorescent algae.
{"title":"Luminescent Microspheres Resolved from Strong Background on an Automated Time-Gated Luminescence Microscopy Workstation","authors":"Len Hamey, R. Connally, Simon Wong Too Yen, Thomas S. Lawson, J. Piper, J. Iredell","doi":"10.1109/DICTA.2009.44","DOIUrl":"https://doi.org/10.1109/DICTA.2009.44","url":null,"abstract":"Fluorescence microscopy is a powerful tool for the rapid identification of target organisms. However, natural autofluorescence often interferes with identification. Time-gated luminescence microscopy (TGLM) uses luminescent labels with long persistence in conjunction with digital imaging to regain discriminative power. Following the excitation pulse, short-lived autofluorescence decays rapidly whereas the long-lived emission from lanthanide doped polymer beads persists for hundreds of microseconds. After a short resolving period, a gated high gain camera captures the persistent emission in the absence of short-lived fluorescence. We report on the development of a TGLM software system for automated scanning of microscope slides, and show its use to resolve luminescent microspheres within a matrix of autofluorescent algae.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122375155","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}
Hand Vein patterns have been adjudged to be one of the safest biometric modalities due to their strong resilience against the impostor attacks. This paper presents a new approach for biometric authentication using infrared thermal hand vein patterns. In contrast to the existing features for hand vein patterns which are based solely on edge detection, we propose Box and branch point based approaches for multiple feature representations. A robust peg free camera set up is employed for infrared thermal imaging. A region of interest (ROI) is extracted from the vein patterns and is convolved with Gabor filter. The real part of this convolution is only preserved for further processing. Multiple features are extracted from the real parts of the convolved images using the proposed branch point based feature extraction techniques. The multiple features are then integrated at the decision level. AND and OR fusion rules are employed to combine the decisions taken by the individual matcher. Experiments conducted on a database of 100 users result in a False Acceptance Rate (FAR) of 0.1% for the Genuine Acceptance Rate (GAR) of 99% for decision level fusion.
{"title":"Biometric Authentication Based on Infrared Thermal Hand Vein Patterns","authors":"Amioy Kumar, M. Hanmandlu, V. Madasu, B. Lovell","doi":"10.1109/DICTA.2009.63","DOIUrl":"https://doi.org/10.1109/DICTA.2009.63","url":null,"abstract":"Hand Vein patterns have been adjudged to be one of the safest biometric modalities due to their strong resilience against the impostor attacks. This paper presents a new approach for biometric authentication using infrared thermal hand vein patterns. In contrast to the existing features for hand vein patterns which are based solely on edge detection, we propose Box and branch point based approaches for multiple feature representations. A robust peg free camera set up is employed for infrared thermal imaging. A region of interest (ROI) is extracted from the vein patterns and is convolved with Gabor filter. The real part of this convolution is only preserved for further processing. Multiple features are extracted from the real parts of the convolved images using the proposed branch point based feature extraction techniques. The multiple features are then integrated at the decision level. AND and OR fusion rules are employed to combine the decisions taken by the individual matcher. Experiments conducted on a database of 100 users result in a False Acceptance Rate (FAR) of 0.1% for the Genuine Acceptance Rate (GAR) of 99% for decision level fusion.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"165 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123330216","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}
Approximate solutions to labelling problems can be found using binary graph cuts and either the alpha-expansion or alpha-beta swap algorithms. In some specific cases, an exact solution can be computed by constructing a multilabel graph. However, in many practical applications the multilabel graph construction is infeasible due to its excessively large memory requirements. In this work, we expand the concept of alpha-beta swap to consider larger sets of labels at each iteration, and demonstrate how this approach is able to produce good approximate solutions to problems which can be solved using multilabel graph cuts. Furthermore, we show how alpha-expansion is a special case of multilabel swap, and from this new formulation, illustrate how alpha-expansion is now able to handle binary energy functions which do not satisfy the triangle inequality. Compared to alpha-beta swap, multilabel swap is able to produce an approximate solution in a shorter amount of time. We demonstrate the merits of our approach by considering the denoising and stereo problems. We illustrate how multilabel swap can be used in a recursive fashion to produce a good solution quickly and without requiring excessive amounts of memory.
{"title":"Solving Multilabel Graph Cut Problems with Multilabel Swap","authors":"Peter Carr, R. Hartley","doi":"10.1109/DICTA.2009.90","DOIUrl":"https://doi.org/10.1109/DICTA.2009.90","url":null,"abstract":"Approximate solutions to labelling problems can be found using binary graph cuts and either the alpha-expansion or alpha-beta swap algorithms. In some specific cases, an exact solution can be computed by constructing a multilabel graph. However, in many practical applications the multilabel graph construction is infeasible due to its excessively large memory requirements. In this work, we expand the concept of alpha-beta swap to consider larger sets of labels at each iteration, and demonstrate how this approach is able to produce good approximate solutions to problems which can be solved using multilabel graph cuts. Furthermore, we show how alpha-expansion is a special case of multilabel swap, and from this new formulation, illustrate how alpha-expansion is now able to handle binary energy functions which do not satisfy the triangle inequality. Compared to alpha-beta swap, multilabel swap is able to produce an approximate solution in a shorter amount of time. We demonstrate the merits of our approach by considering the denoising and stereo problems. We illustrate how multilabel swap can be used in a recursive fashion to produce a good solution quickly and without requiring excessive amounts of memory.","PeriodicalId":277395,"journal":{"name":"2009 Digital Image Computing: Techniques and Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129227866","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：