{"title":"A Measure of Perceptual Aliasing in Image Descriptors","authors":"Jie Wang, Hong Zhang","doi":"10.1109/CRV.2013.27","DOIUrl":"https://doi.org/10.1109/CRV.2013.27","url":null,"abstract":"Researchers exploring problems in image matching tasks face the curse of perceptual aliasing that is originally used in characterizing a sensing process. Perceptual aliasing occurs when the one-to-one mapping relations between world states (objects) and their representations (descriptors) are not maintained. In this paper, we introduce a novel method for quantifying perceptual aliasing. Our method measures the discriminating power of an image descriptor in terms of its ability to distinguish between images of different objects and to match images of the same object. To illustrate our method, we apply it in the evaluation of popular global image descriptors that do not require local feature or keypoint detection. Specifically, our method runs spectral clustering on the similarity matrix computed with descriptors of known image clusters and measures the performance of an image descriptor by its ability to maintain the original clusters, using two indices, MRI-1 and MRI-2, that are based on Rand index.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123776555","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}
Laszlo-Peter Berczi, J. Gammell, Chi Hay Tong, M. Daly, T. Barfoot
{"title":"A Proof-of-Concept, Rover-Based System for Autonomously Locating Methane Gas Sources on Mars","authors":"Laszlo-Peter Berczi, J. Gammell, Chi Hay Tong, M. Daly, T. Barfoot","doi":"10.1109/CRV.2013.16","DOIUrl":"https://doi.org/10.1109/CRV.2013.16","url":null,"abstract":"This paper presents a proof-of-concept, rover-based system to locate the source(s) of methane gas on Mars. A distributed open-path spectrometer is mounted on a rover and pointed at several retroreflective signs to measure the line-of-sight methane concentration between the rover and sign. By moving the rover around and accumulating such measurements, the location of a methane source can be determined. The system was integrated and tested over a 2 kilometre traverse in an analog mission at the Canadian Space Agency on the Mars Emulation Terrain. Engineering results suggest that total driving distance is comparable to other similar methods, but by using a network of reusable paths, the total distance driven in new terrain can be reduced by a factor of five. Autonomous pointing of signs was successful 80% of the time. Future work includes combining the engineering and science data from the analog mission to better understand how to locate methane sources.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116540383","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}
{"title":"OpenVL: Abstracting Vision Tasks Using a Segment-Based Language Model","authors":"G. Miller, S. Oldridge, S. Fels","doi":"10.1109/CRV.2013.55","DOIUrl":"https://doi.org/10.1109/CRV.2013.55","url":null,"abstract":"Computer vision is a complex field which can be challenging for those outside the community to apply in the real world. In this paper we show one method to provide access to sophisticated computer vision methods to general developers, hobbyists or researchers outside the field. Our contribution is an abstraction utilising fundamental vision operations based on a single unit, the segment, can be used to describe images, local image conditions and between-image conditions. We illustrate how a descriptive language model can be built on the segment to provide an intuitive mental model of computer vision to mainstream developers. We demonstrate how we can map a description of the task composed of the segment-based language into the space of algorithms, to choose an appropriate method to solve the problem. We use the problems of segmentation, correspondence and image registration to show how end-to-end problems may be constructed using our novel metaphor.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"318 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116286786","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}
{"title":"Efficient, General Point Cloud Registration with Kernel Feature Maps","authors":"Hanchen Xiong, S. Szedmák, J. Piater","doi":"10.1109/CRV.2013.26","DOIUrl":"https://doi.org/10.1109/CRV.2013.26","url":null,"abstract":"This paper proposes a novel and efficient point cloud registration algorithm based on the kernel-induced feature map. Point clouds are mapped to a high-dimensional (Hilbert) feature space, where they are modeled with Gaussian distributions. A rigid transformation is first computed in feature space by elegantly computing and aligning a small number of eigenvectors with kernel PCA (KPCA) and is then projected back to 3D space by minimizing a consistency error. SE(3) on-manifold optimization is employed to search for the optimal rotation and translation. This is very efficient; once the object-specific eigenvectors have been computed, registration is performed in linear time. Because of the generality of KPCA and SE(N) on-manifold method, the proposed algorithm can be easily extended to registration in any number of dimensions (although we only focus on 3D case). The experimental results show that the proposed algorithm is comparably accurate but much faster than state-of-the-art methods in various challenging registration tasks.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134117864","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}
{"title":"Exploiting Color Constancy for Compensating Projected Images on Non-white Light Projection Screen","authors":"A. Madi, D. Ziou, Frederic Dhalleine","doi":"10.1109/CRV.2013.49","DOIUrl":"https://doi.org/10.1109/CRV.2013.49","url":null,"abstract":"The use of projectors in various applications and environments has raised the need to adapt the displayed image to the viewing display environment. The displayed image under bright light may appear with washed colors and attenuated contrast. Similarly, the appearance of the displayed image may be affected by the reflectance and the geometry of the projection screen. In this paper, we formulate the problem of the color correction using the model of the perceived image, by exploiting the color constancy concept. We demonstrate that the color constancy can be widely employed to make perceived images invariant to the viewing conditions and even to simulate appearances obtained in some desired environments. We verify through formal and empirical validations that the proposed methods significantly produce the desired colors. Our methods can be useful in some areas like camouflages and simulations.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134195849","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}
{"title":"Tracking Severe Storms Using a Pseudo Storm Concept","authors":"Yong Zhang, Robert E. Mercer, J. Barron, P. Joe","doi":"10.1109/CRV.2013.9","DOIUrl":"https://doi.org/10.1109/CRV.2013.9","url":null,"abstract":"Tracking storms in radar images can be conceived of as a problem of tracking deformable objects. Our current relaxation labelling-based tracking algorithm that represents these deformable objects as “fuzzy” points can track objects that undergo shape deformations. One other type of deformation is the splitting of an object into multiple objects or the merging of multiple objects into one object from one image to the next. With our current algorithm, tracks are interrupted when such events happen in image sequences. We remove this deficiency of the current algorithm by adding the concept of a Pseudo Storm to its representational repertoire. With only minor modifications to the current algorithm, the new algorithm can track deformable objects that undergo both merging and splitting events. The new pseudo storm tracking algorithm outperforms our previous storm tracking algorithm for Great Lakes Doppler precipitation datasets.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"113 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124158332","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}
{"title":"3D Disaster Scene Reconstruction Using a Canine-Mounted RGB-D Sensor","authors":"Jimmy Tran, Alex Ufkes, A. Ferworn, M. Fiala","doi":"10.1109/CRV.2013.15","DOIUrl":"https://doi.org/10.1109/CRV.2013.15","url":null,"abstract":"A 3D map of the interior of a disaster site that pinpoints the location of trapped victims would greatly aid search and rescue efforts. We propose using a canine-mounted RGB-D sensor; a trained rescue dog can carry an image sensor through the site to build a 3D model useful for rescuers. However, the registration of the data provides challenges beyond those typically faced in scene reconstruction due to the rapid motion and sudden pose changes. We provide a solution whereby a pre-processing step identifies good frames to combine from a stream of RGB-D image frames. These selected images are then combined into the larger model by calculating a relative pose using the 3D location of key points matched in the visible images. Results are presented of 3D models constructed using data collected from the canine platform.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124651353","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}
{"title":"I Remember Seeing This Video: Image Driven Search in Video Collections","authors":"Zheng Wang, F. Qureshi","doi":"10.1109/CRV.2013.31","DOIUrl":"https://doi.org/10.1109/CRV.2013.31","url":null,"abstract":"We present a novel technique for image driven shot retrieval in video data. Specifically, given a query image, our method can efficiently pick the video segment containing that image. Video is first divided into shots. Each shot is described using an embedded hidden Markov model (EHMM). The EHMM is trained on GIST-like descriptors of frames in that shot. The trained EHMM computes the likelihood that a query image belongs to the shot. A Support Vector Machine classifier is trained for each EHMM. The classifier provides a yes/no decision given the likelihood value produced by its EHMM. Given a collection of shot models from one or more videos, the proposed technique can efficiently decide whether or not an image belongs to a video by identifying the shot most likely to contain that image. The proposed technique is evaluated on a realistic dataset.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133250928","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}
J. I. Vasquez-Gomez, L. Sucar, Rafael Murrieta-Cid
{"title":"Hierarchical Ray Tracing for Fast Volumetric Next-Best-View Planning","authors":"J. I. Vasquez-Gomez, L. Sucar, Rafael Murrieta-Cid","doi":"10.1109/CRV.2013.42","DOIUrl":"https://doi.org/10.1109/CRV.2013.42","url":null,"abstract":"A mobile robot must have the ability of building a representation of its environment and the objects in it. To build a three-dimensional (3D) model of a physical object, several scans must be taken at different locations. Selecting each location is the next-best-view problem. Search based methods, where candidate views are generated and evaluated by a utility function, are a solution. However, such methods are slow for high resolution models given that the evaluation requires visibility computation in 3D. We propose a scene representation by octrees with a hierarchical ray tracing that reduces the visibility computation time. Such method performs a coarse ray tracing, except for the interesting volumes where a finer resolution is applied. The method decreases the computation time at least one order of magnitude. Saving time with this method leads to evaluate more constraints and more candidate views in high resolution models.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"95 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124737320","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}
{"title":"Parallel Planes Identification Using Uncalibrated Zooming Cameras","authors":"Tarik Elamsy, B. Boufama, A. Habed","doi":"10.1109/CRV.2013.41","DOIUrl":"https://doi.org/10.1109/CRV.2013.41","url":null,"abstract":"This paper proposes a new method for the automatic identification of parallel planes form uncalibrated images of a scene. First, we demonstrate that, given a priori knowledge about a single pair of parallel planes, it is possible to identify all other parallel planes from uncalibrated images of a scene. Then, we show that a pair of parallel virtual planes can be made available when the scene is imaged by stationary non-rotating zooming cameras leading to a fully automatic identification method.Experiments with indoor and outdoor real images have validated our method.","PeriodicalId":132664,"journal":{"name":"2013 International Conference on Computer and Robot Vision","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129116833","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}