Modeling and classifying symmetries using a multiscale opponent color representation

Thai, Bea; Healey, Glenn

doi:10.1109/34.730556

Cited by 30 publications

(13 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We use a bank of Gabor filters in a configuration known as multichannel filtering for texture characterization and classification. The strength of multichannel Gabor filtering and the advantages of its application in the field of image processing and pattern recognition have been confirmed by many researchers and have received considerable attention in the vision community [5], [14]- [17]. The underlying reason for this strong support is based on evidence that the Gabor spectrum and representation is computed in the visual cortex of mammals [12], [13].…”

Section: Gabor Channel Filtersmentioning

confidence: 93%

Multiscale color invariants based on the human visual system

Wanderley

Fisher

2001

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

This paper proposes a new representation for color texture using a set of multiscale illuminant invariant features. The approach was specifically developed to investigate the feasibility of using machine vision to automatically monitor populations of animal species in ecologically sensitive regions, such as the Amazon Forest. The approach uses a combination of Finlayson's (1994) color angle idea and Gabor multichannel filters and was inspired by the multichannel model of the human visual system (HVS). Using a database of color textures from three species of Amazonian monkey, and also a previously published reference database of color regions, we show that the approach performs better than methods based on color angles or Gabor filters alone. The Monkey database was compiled from texture segments extracted from a video of the Amazon Forest using a spatio-temporal segmentation algorithm. The approach is evaluated by applying two different classification tests in order to measure the quality of the recognition features root mean square (RMS) analysis and receiver operating characteristic (ROC) analysis.

show abstract

Section: Gabor Channel Filtersmentioning

confidence: 93%

Multiscale color invariants based on the human visual system

Wanderley

Fisher

2001

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

show abstract

“…Although there have been fewer attempts to integrate texture and chrominance information [25], color and texture have been segmented separately, then fused in a postprocessing step that typically yields greater accuracy than either the color or texture result alone [26]. For example, each spectral band can be processed separately using texture analysis techniques derived from (a) monochromatic image analysis [27,28]; from luminance, supported by pure chrominance information [29,30], or from chrominance bands via inter-band correlation [31][32][33][34]. Muñoz, et al published a review of, and novel techniques for, segmentation of multispectral textured images, identifying seven strategies for integration of color and texture [35], including fusion of region and boundary information [36].…”

Section: Color and Texture Segmentationmentioning

confidence: 99%

<title>Error measures for object-based image compression</title>

Schmalz

Ritter

2005

Mathematics of Data/Image Coding, Compression, and Encryption VIII, With Applications

View full text Add to dashboard Cite

Image compression based on transform coding appears to be approaching a bit-rate limit for visually acceptable distortion levels. Although an emerging compression technology called object-based compression (OBC) promises significantly improved bit rate and computational efficiency, OBC is epistemologically distinct in a way that renders existing image quality measures (IQMs) for compression transform optimization less suitable for OBC. In particular, OBC segments source image regions, then efficiently encodes each region's content and boundary. During decompression, region contents are often replaced by similar-appearing objects from a codebook, thus producing a reconstructed image that corresponds semantically to the source image, but has pixel-, featural-, and object-level differences that are apparent visually. OBC thus gains the advantage of fast decompression via efficient codebook-based substitutions, albeit at the cost of codebook search in the compression step and significant pixel-or region-level errors in decompression. Existing IQMs are pixel-and regionoriented, and thus tend to indicate high error due to OBC's lack of pixel-level correlation between source and reconstructed imagery. Thus, current IQMs do not necessarily measure the semantic correspondence that OBC is designed to produce. This paper presents image quality measures for estimating semantic correspondence between a source image and a corresponding OBC-decompressed image. In particular, we examine the semantic assumptions and models that underlie various approaches to OBC, especially those based on textural as well as high-level name and spatial similarities. We propose several measures that are designed to quantify this type of high-level similarity, and can be combined with existing IQMs for assessing compression transform performance. Discussion also highlights how these novel IQMs can be combined with time and space complexity measures for compression transform optimization.

show abstract

“…For example, each spectral band can be processed separately using texture analysis techniques derived from monochromatic image analysis [22,23]. Alternatively, textural information can be derived from luminance, supported by pure chrominance information [24,25].…”

Section: Joint Color-texture Segmentationmentioning

confidence: 99%

<title>Computational complexity of object-based image compression</title>

Schmalz

Ritter

2005

Mathematics of Data/Image Coding, Compression, and Encryption VIII, With Applications

View full text Add to dashboard Cite

Image compression via transform coding applied to small rectangular regions or encoding blocks appears to be approaching asymptotic rate-distortion performance. However, an emerging compression technology, called object-based compression (OBC) promises significantly improved performance via compression ratios ranging from 200:1 to as high as 2,500:1. OBC involves segmentation of image regions, followed by efficient encoding of each region's content and boundary. During decompression, such regions can be approximated by objects from a codebook, yielding a reconstructed image that is semantically equivalent to the corresponding source image, but has pixel-and featural-level differences.Semantic equivalence between the source and decompressed image facilitates fast decompression through efficient substitutions, albeit at the cost of codebook search in the compression step. Given small codebooks, OBC holds promise for information-push technologies where approximate context is sufficient, for example, transmission of surveillance images that provide the gist of a scene. However, OBC is not necessarily useful for applications requiring high accuracy, such as medical image processing, because substitution of source content can be inaccurate at small spatial scales.The cost of segmentation is a significant disadvantage in current OBC implementations. Several innovative techniques have been developed for region segmentation, as discussed in a previous paper [4]. Additionally, tradeoffs between representational fidelity, computational cost, and storage requirement occur, as with the vast majority of lossy compression algorithms. This paper analyzes the computational (time) and storage (space) complexities of several recent OBC algorithms applied to single-frame imagery. A time complexity model is proposed, which can be associated theoretically with a space complexity model that we have previously published [2]. The result, when combined with measurements of representational accuracy described in a companion paper [5], supports estimation of a time-space-error bandwidth product that could facilitate dynamic optimization of OBC algorithms. In practice, this would support efficient compression with visually acceptable reconstruction for a wide variety of military and domestic applications.

show abstract

Modeling and classifying symmetries using a multiscale opponent color representation

Cited by 30 publications

References 24 publications

Multiscale color invariants based on the human visual system

Multiscale color invariants based on the human visual system

<title>Error measures for object-based image compression</title>

<title>Computational complexity of object-based image compression</title>

Contact Info

Product

Resources

About