In oblique shape from shading (SfS), the lighting (illumination) direction is essential for recovering the 3D surface of a shaded image. On the other hand, fast marching methods (FMM) are SfS algorithms that use the mechanism of wave propagation to reconstruct the surface. In this paper, the estimation of lighting direction is addressed and we model it as an optimization problem. The idea is to minimize the inconsistency of wave propagation of FMM during the reconstruction. As the consistency of wave propagation is a multi-modal function, genetic algorithm (GA) is utilized to determine the lighting direction. Experimental results on four oblique SfSs with an unknown lighting direction are presented to demonstrate the performance of the proposed algorithm.
{"title":"Lighting Direction Estimation in Perspective Shape from Shading by Genetic Algorithm","authors":"C. Chow, S. Y. Yuen","doi":"10.1109/CRV.2007.43","DOIUrl":"https://doi.org/10.1109/CRV.2007.43","url":null,"abstract":"In oblique shape from shading (SfS), the lighting (illumination) direction is essential for recovering the 3D surface of a shaded image. On the other hand, fast marching methods (FMM) are SfS algorithms that use the mechanism of wave propagation to reconstruct the surface. In this paper, the estimation of lighting direction is addressed and we model it as an optimization problem. The idea is to minimize the inconsistency of wave propagation of FMM during the reconstruction. As the consistency of wave propagation is a multi-modal function, genetic algorithm (GA) is utilized to determine the lighting direction. Experimental results on four oblique SfSs with an unknown lighting direction are presented to demonstrate the performance of the proposed algorithm.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131964436","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}
This paper presents a hybrid method for the segmentation of SAR sea ice images, which consists of an initial watershed segmentation followed by a region merging. Iterative bilateral filtering is used to reduce speckle noise and suppress irrelevant image details, which can significantly alleviate oversegmentation of watersheds. Since edges are well preserved by bilateral filtering, the watershed algorithm is capable of precisely locating object boundaries. Final segmentation is accomplished by applying an iterative region merging on the watershed regions by taking into account local boundary strengths and regional statistics. The efficiency of the proposed method has been demonstrated on the segmentation of SAR sea ice images. In comparison with traditional watershed algorithm, our method achieves better performance in identifying filament structures such as leads.
{"title":"SAR Sea Ice Image Segmentation Based on Edge-preserving Watersheds","authors":"Xuezhi Yang, David A Clausi","doi":"10.1109/CRV.2007.58","DOIUrl":"https://doi.org/10.1109/CRV.2007.58","url":null,"abstract":"This paper presents a hybrid method for the segmentation of SAR sea ice images, which consists of an initial watershed segmentation followed by a region merging. Iterative bilateral filtering is used to reduce speckle noise and suppress irrelevant image details, which can significantly alleviate oversegmentation of watersheds. Since edges are well preserved by bilateral filtering, the watershed algorithm is capable of precisely locating object boundaries. Final segmentation is accomplished by applying an iterative region merging on the watershed regions by taking into account local boundary strengths and regional statistics. The efficiency of the proposed method has been demonstrated on the segmentation of SAR sea ice images. In comparison with traditional watershed algorithm, our method achieves better performance in identifying filament structures such as leads.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129771937","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 present a performance evaluation framework for visual feature extraction and matching in the visual simultaneous localization and mapping (SLAM) context. Although feature extraction is a crucial component, no qualitative study comparing different techniques from the visual SLAM perspective exists. We extend previous image pair evaluation methods to handle non-planar scenes and the multiple image sequence requirements of our application, and compare three popular feature extractors used in visual SLAM: the Harris corner detector, the Kanade-Lucas-Tomasi tracker (KLT), and the scale-invariant feature transform (SIFT). We present results from a typical indoor environment in the form of recall/precision curves, and also investigate the effect of increasing distance between image viewpoints on extractor performance. Our results show that all methods can be made to perform well, although it is possible to distinguish between the three. We conclude by presenting guidelines for selecting a feature extractor for visual SLAM based on our experiments.
{"title":"Quantitative Evaluation of Feature Extractors for Visual SLAM","authors":"J. Klippenstein, Hong Zhang","doi":"10.1109/CRV.2007.52","DOIUrl":"https://doi.org/10.1109/CRV.2007.52","url":null,"abstract":"We present a performance evaluation framework for visual feature extraction and matching in the visual simultaneous localization and mapping (SLAM) context. Although feature extraction is a crucial component, no qualitative study comparing different techniques from the visual SLAM perspective exists. We extend previous image pair evaluation methods to handle non-planar scenes and the multiple image sequence requirements of our application, and compare three popular feature extractors used in visual SLAM: the Harris corner detector, the Kanade-Lucas-Tomasi tracker (KLT), and the scale-invariant feature transform (SIFT). We present results from a typical indoor environment in the form of recall/precision curves, and also investigate the effect of increasing distance between image viewpoints on extractor performance. Our results show that all methods can be made to perform well, although it is possible to distinguish between the three. We conclude by presenting guidelines for selecting a feature extractor for visual SLAM based on our experiments.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130962839","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}
This paper presents a simple linear operator that accurately estimates the position and parameters of ellipse features. Based on the dual conic model, the operator avoids the intermediate stage of precisely extracting individual edge points by exploiting directly the raw gradient information in the neighborhood of an ellipse's boundary. Moreover, under the dual representation, the dual conic can easily be constrained to a dual ellipse when minimizing the algebraic distance. The new operator is assessed and compared to other estimation approaches in simulation as well as in real situation experiments and shows better accuracy than the best approaches, including those limited to the center position.
{"title":"A Simple Operator for Very Precise Estimation of Ellipses","authors":"Jean-Nicolas Ouellet, P. Hébert","doi":"10.1109/CRV.2007.8","DOIUrl":"https://doi.org/10.1109/CRV.2007.8","url":null,"abstract":"This paper presents a simple linear operator that accurately estimates the position and parameters of ellipse features. Based on the dual conic model, the operator avoids the intermediate stage of precisely extracting individual edge points by exploiting directly the raw gradient information in the neighborhood of an ellipse's boundary. Moreover, under the dual representation, the dual conic can easily be constrained to a dual ellipse when minimizing the algebraic distance. The new operator is assessed and compared to other estimation approaches in simulation as well as in real situation experiments and shows better accuracy than the best approaches, including those limited to the center position.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115046199","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 describe a video indexing system that aims at indexing large video files in relation to the presence of similar faces. The detection of near-frontal view faces is done with a cascade of weak classifier. Face tracking is done through a particle filter and generate trajectories. Face clusters are found based on a spectral clustering approach. We compare the performance of various spectral clustering techniques based on 2DPCA features. The system performance is evaluated against a public face database as well as on a real full-length feature movie.
{"title":"Automatic Detection and Clustering of Actor Faces based on Spectral Clustering Techniques","authors":"S. Foucher, L. Gagnon","doi":"10.1109/CRV.2007.13","DOIUrl":"https://doi.org/10.1109/CRV.2007.13","url":null,"abstract":"We describe a video indexing system that aims at indexing large video files in relation to the presence of similar faces. The detection of near-frontal view faces is done with a cascade of weak classifier. Face tracking is done through a particle filter and generate trajectories. Face clusters are found based on a spectral clustering approach. We compare the performance of various spectral clustering techniques based on 2DPCA features. The system performance is evaluated against a public face database as well as on a real full-length feature movie.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132267938","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}
Direct methods of image registration work by defining a measure of the difference between two images and using numerical optimization methods to find the transformation that minimizes the difference. It has often been proposed that these methods may be speeded up by using only a sub- set of pixels to compute the difference measure. Previous work has suggested some criteria to use in pixel selection based on the derivative of the image, but has not addressed the issue of performance degradation that can result from applying these techniques. In this paper, we show that un- less applied carefully, these methods do not actually help. Specifically, reliability of the registration algorithm is lost if the initial starting position is further from the optimum than the scale of the derivative. Additionally, we propose new criteria for pixel selection which are strongly based on in- formation theory, and are faster to compute. We verify these propositions for two popular image difference measures by examining their behavior as the transformation parameters are varied, and by registering a number of typical images.
{"title":"The importance of scale when selecting pixels for image registration","authors":"Rupert Brooks, T. Arbel","doi":"10.1109/CRV.2007.64","DOIUrl":"https://doi.org/10.1109/CRV.2007.64","url":null,"abstract":"Direct methods of image registration work by defining a measure of the difference between two images and using numerical optimization methods to find the transformation that minimizes the difference. It has often been proposed that these methods may be speeded up by using only a sub- set of pixels to compute the difference measure. Previous work has suggested some criteria to use in pixel selection based on the derivative of the image, but has not addressed the issue of performance degradation that can result from applying these techniques. In this paper, we show that un- less applied carefully, these methods do not actually help. Specifically, reliability of the registration algorithm is lost if the initial starting position is further from the optimum than the scale of the derivative. Additionally, we propose new criteria for pixel selection which are strongly based on in- formation theory, and are faster to compute. We verify these propositions for two popular image difference measures by examining their behavior as the transformation parameters are varied, and by registering a number of typical images.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127828382","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}
Idalia Maldonado-Castillo, M. Eramian, R. Pierson, J. Singh, G. Adams
Studies of ovarian development in female mammals have shown a relationship between the day in the estrous cycle and the size of the main structures and physiological status of the ovary. This paper presents an algorithm for the automatic classification of bovine ovaries into temporal categories using information extracted from ultrasound images. The temporal classes corresponded roughly to the metestrus, diestrus, and proestrus phases of the bovine reproductive cycle. Features based on the sizes of ovarian structures formed the patterns on which the classification was performed. A Naive Bayes classifier was able to correctly classify the stage of the estrous cycle for 86.36% of the test patterns. A decision tree classified 100% of the test patterns correctly. The decision tree inference algorithm used to build the classifier constructed a tree that used only two of the five available features indicating that they form a sufficiently rich set of features for robust classification.
{"title":"Classification of Bovine Reproductive Cycle Phase using Ultrasound-Detected Features","authors":"Idalia Maldonado-Castillo, M. Eramian, R. Pierson, J. Singh, G. Adams","doi":"10.1109/CRV.2007.16","DOIUrl":"https://doi.org/10.1109/CRV.2007.16","url":null,"abstract":"Studies of ovarian development in female mammals have shown a relationship between the day in the estrous cycle and the size of the main structures and physiological status of the ovary. This paper presents an algorithm for the automatic classification of bovine ovaries into temporal categories using information extracted from ultrasound images. The temporal classes corresponded roughly to the metestrus, diestrus, and proestrus phases of the bovine reproductive cycle. Features based on the sizes of ovarian structures formed the patterns on which the classification was performed. A Naive Bayes classifier was able to correctly classify the stage of the estrous cycle for 86.36% of the test patterns. A decision tree classified 100% of the test patterns correctly. The decision tree inference algorithm used to build the classifier constructed a tree that used only two of the five available features indicating that they form a sufficiently rich set of features for robust classification.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127609278","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 a wavelet-based light representation. This representation is used for recovering illumination and reflectance of a scene with known geometry from a single calibrated image. Previous approaches of light estimation reconstruct either a discrete set of infinite light sources or the projection of light on the spherical harmonics basis. The first approach is more suitable for modeling sharp light effects while the second one works best for diffuse scenes. In contrast, we show that the proposed representation is suitable for both diffuse and specular scenes. We compared our technique with the two previously mentioned approaches and found it superior. In addition to illumination estimation our algorithm also estimates a Phong surface reflection. Experiments with synthetic and real scenes demonstrate the effectiveness of our method.
{"title":"Wavelet-based Light Reconstruction from a Single Image","authors":"Cameron Upright, Dana Cobzas, Martin Jägersand","doi":"10.1109/CRV.2007.70","DOIUrl":"https://doi.org/10.1109/CRV.2007.70","url":null,"abstract":"In this paper we propose a wavelet-based light representation. This representation is used for recovering illumination and reflectance of a scene with known geometry from a single calibrated image. Previous approaches of light estimation reconstruct either a discrete set of infinite light sources or the projection of light on the spherical harmonics basis. The first approach is more suitable for modeling sharp light effects while the second one works best for diffuse scenes. In contrast, we show that the proposed representation is suitable for both diffuse and specular scenes. We compared our technique with the two previously mentioned approaches and found it superior. In addition to illumination estimation our algorithm also estimates a Phong surface reflection. Experiments with synthetic and real scenes demonstrate the effectiveness of our method.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122488124","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}
This paper explores the possibility of assessing the adequacy of a training database to be used in a learning-based super-resolution process. The Mean Euclidean Distance (MED) function is obtained by averaging the distance between each input patch and its closest candidate in the training database, for a series of blurring kernels used to construct the low-resolution database. The shape of that function is thought to indicate the level of adequacy of the database, thus indicating to the user the potential of success of a learning-based super-resolution algorithm using this database.
{"title":"Training Database Adequacy Analysis for Learning-Based Super-Resolution","authors":"I. Bégin, F. Ferrie","doi":"10.1109/CRV.2007.65","DOIUrl":"https://doi.org/10.1109/CRV.2007.65","url":null,"abstract":"This paper explores the possibility of assessing the adequacy of a training database to be used in a learning-based super-resolution process. The Mean Euclidean Distance (MED) function is obtained by averaging the distance between each input patch and its closest candidate in the training database, for a series of blurring kernels used to construct the low-resolution database. The shape of that function is thought to indicate the level of adequacy of the database, thus indicating to the user the potential of success of a learning-based super-resolution algorithm using this database.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117353215","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}
An increasingly popular approach to support military forces deployed in urban environments consists in using autonomous robots to carry on critical tasks such as mapping and surveillance. In order to cope with the complex obstacles and structures found in this operational context, robots should be able to perceive and analyze their world in 3D. The method presented in this paper uses a 3D volumetric sensor to efficiency map and explore urban environments with an autonomous robotic platform. A key feature of our work is that the 3D model of the environment is preserved all along the process using a multiresolution octree. This way, every module can access the information it contains to achieve its tasks. Simulation and real word tests were performed to validate the performance of the integrated system and are presented at the end of the paper.
{"title":"Mapping and Exploration of Complex Environments Using Persistent 3D Model","authors":"J. Fournier, B. Ricard, D. Laurendeau","doi":"10.1109/CRV.2007.45","DOIUrl":"https://doi.org/10.1109/CRV.2007.45","url":null,"abstract":"An increasingly popular approach to support military forces deployed in urban environments consists in using autonomous robots to carry on critical tasks such as mapping and surveillance. In order to cope with the complex obstacles and structures found in this operational context, robots should be able to perceive and analyze their world in 3D. The method presented in this paper uses a 3D volumetric sensor to efficiency map and explore urban environments with an autonomous robotic platform. A key feature of our work is that the 3D model of the environment is preserved all along the process using a multiresolution octree. This way, every module can access the information it contains to achieve its tasks. Simulation and real word tests were performed to validate the performance of the integrated system and are presented at the end of the paper.","PeriodicalId":304254,"journal":{"name":"Fourth Canadian Conference on Computer and Robot Vision (CRV '07)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115525678","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：