Retrieval by classification of images containing large manmade objects using perceptual grouping

Qasim, Iqbal; Aggarwal, J. K.

doi:10.1016/s0031-3203(01)00139-x

Cited by 79 publications

(42 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Iqbal and Aggarwal (2002) proposed to use the perceptual grouping of edge features for classification. The main idea here is to count evidences of the target class in the image.…”

Section: State Of the Artmentioning

confidence: 99%

“…The more expressive the primitive is, the smaller the feature vector is, but more difficult the extraction becomes. To compare with Iqbal and Aggarwal (2002), circular primitives are also extracted, and a larger feature vector, taking into consideration primitives' geometric, relational and spatial attributes, is constructed.…”

Section: The Primitive Extraction and Representationmentioning

confidence: 99%

See 1 more Smart Citation

How to involve structural modeling for cartographic object recognition tasks in high-resolution satellite images?

Erus

Loménie

2010

Pattern Recognition Letters

View full text Add to dashboard Cite

To cite this version:Guray Erus, Nicolas Lomenie.How to involve structural modeling for cartographic object recognition tasks in high-resolution satellite images?.Elsevier AbstractWith the new generation of satellite systems, very high resolution satellite images will be available daily at a high delivery rate. The exploitation of such a huge amount of data will be made possible by the design of high performance analysis algorithms for decision making systems. In particular, the detection and recognition of complex man-made objects is a new challenge coming with this new level of resolution. In this study, we develop a system that recognizes such structured and compact objects like bridges or roundabouts. The original contribution of this work is the use of structural shape attributes in an appearance based statistical learning method framework leading to valuable recognition and false alarm rates. This hybrid structural/statistical approach aims to construct an intermediate step between the low-level image characteristics and high-level semantic concepts.

show abstract

“…Iqbal and Aggarwal (2002) proposed to use the perceptual grouping of edge features for classification. The main idea here is to count evidences of the target class in the image.…”

Section: State Of the Artmentioning

confidence: 99%

Section: The Primitive Extraction and Representationmentioning

confidence: 99%

How to involve structural modeling for cartographic object recognition tasks in high-resolution satellite images?

Erus

Loménie

2010

Pattern Recognition Letters

View full text Add to dashboard Cite

show abstract

“…The di culties with the segmentation-based approach have driven many researchers to use a low-level feature, exemplar-based approach (e.g., [5][6][7][8][9][10][11][12][13][14][15][16][17][18]). While many have taken this approach, none handle the multi-label problem.…”

Section: Related Workmentioning

confidence: 99%

“…A problem domain receiving renewed attention is semantic scene classiÿcation [5][6][7][8][9][10][11][12][13][14][15][16][17][18], categorizing images into semantic classes such as beaches, sunsets or parties. Semantic scene classiÿcation ÿnds application in many areas, including content-based indexing and organization and content-sensitive image enhancement.…”

Section: Introductionmentioning

confidence: 99%

Learning multi-label scene classification*1

Boutell

2004

Pattern Recognition

View full text Add to dashboard Cite

In classic pattern recognition problems, classes are mutually exclusive by deÿnition. Classiÿcation errors occur when the classes overlap in the feature space. We examine a di erent situation, occurring when the classes are, by deÿnition, not mutually exclusive. Such problems arise in semantic scene and document classiÿcation and in medical diagnosis. We present a framework to handle such problems and apply it to the problem of semantic scene classiÿcation, where a natural scene may contain multiple objects such that the scene can be described by multiple class labels (e.g., a ÿeld scene with a mountain in the background). Such a problem poses challenges to the classic pattern recognition paradigm and demands a di erent treatment. We discuss approaches for training and testing in this scenario and introduce new metrics for evaluating individual examples, class recall and precision, and overall accuracy. Experiments show that our methods are suitable for scene classiÿcation; furthermore, our work appears to generalize to other classiÿcation problems of the same nature.

show abstract

“…The latest implementation of the system [11][12][13] is able to serve queries ranging from scenes of purely natural objects such as vegetation, trees and sky, to images containing conspicuous structural objects such as buildings, towers and bridges. The system was tested on an image database consisting of 2660 24-bit color images.…”

Section: D Reconstruction Of An Urban Scene From Synthetic Fish-eve mentioning

confidence: 99%

A Research Center in the Scientific Foundation of Image Representation and Analysis (CIS)

Miller¹

2001

PsycEXTRA Dataset

View full text Add to dashboard Cite

Public Reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comment regarding this burden estimates or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and The views, opinions and/or findings contained in this teport are thcJ&e^pf the authors) and should not be construed as an official Department of the Army position, policy or decision, unless so designated by other documentation. The efforts of CIS during the 1999-2001 years built on the mathematical foundations that have been emerging over the past several decades. Examining a broad class of remote sensing problems, we have been establishing the fundamental framework for the inference and representation of structures in complex systems and scenes of complex shapes proceeding from the representation of complex scenes, to image formation and sensor modeling, and culminating in the development of the fundamental underpinnings for optimal decision/recognition strategies in image understanding and ATR. Within this framework, we are establishing the methodology for establishing the limits of performance of detection, identification and recognition algorithms solving remote sensing problems involving data from multiple active and passive sensors. The Web site continues to receive 5000-7000 visitors a month and has distributed databases to an average of 500 unique visitors each month. CIS members have submitted for publication 6 papers, published 34 papers, gave 32 talks, and have been involved in numerous meetings with the Army and DOD. CIS has a publications index of 253 papers in the areas of target recognition, deformable geometry, image formation and sensor modeling, tomographic imaging, computational vision, inference and optimization, adaptation and learning, multiple sensor fusion, clutter modeling, and performance bounds. SUBJECT TERMS Automatic Target Recognition Summary of the most important resultsThe major results in the Center: 1. Performance bounds and metrics: We have established Performance Bounds and Metrics for pose estimation and identification. We have established for pose the MMSE estimator, and established curves of performances for multiple sensors, FLIR, LADAR, HRR, VIDEO.We have established the tight connection between estimator bounds and ID bounds. Exponential ID bounds are determined by estimator accuracy bounds. Sensor fusion:We have established the optimum methodology for combining information optimally from multiple sensors. Estimation and ID bounds have been calculated for combinations of HRR, LADAR, and FLIR.3. FLIR signature variation: We have established a methodology for accommodating the infinit...

show abstract

Retrieval by classification of images containing large manmade objects using perceptual grouping

Cited by 79 publications

References 34 publications

How to involve structural modeling for cartographic object recognition tasks in high-resolution satellite images?

How to involve structural modeling for cartographic object recognition tasks in high-resolution satellite images?

Learning multi-label scene classification*1

A Research Center in the Scientific Foundation of Image Representation and Analysis (CIS)

Contact Info

Product

Resources

About