Illumination Invariance and Object Model in Content-Based Image and Video Retrieval

Li, Ze-Nian; Zaı̈ane, Osmar R.; Tauber, Zinovi

doi:10.1006/jvci.1998.0403

Cited by 44 publications

(35 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Queries typically comprise blobs of color or predefined texture [11,15]. Color and texture information is augmented with shape descriptors in [18], and spectral descriptors such as wavelets have also been explored [12].…”

Section: Related Workmentioning

confidence: 99%

Reverse storyboarding for video retrieval

McNeill

Collomosse

2007

IET 4th European Conference on Visual Media Production (CVMP 2007)

View full text Add to dashboard Cite

We present a novel Content Based Video Retrieval (CBVR) system, driven by free-hand sketch queries depicting both objects and their movement (via dynamic cues; streak-lines and arrows). Our main contribution is a probabilistic model of video clips (based on Linear Dynamical Systems), leading to an algorithm for matching descriptions of sketched objects to video. We demonstrate our model fitting to clips under static and moving camera conditions, exhibiting linear and oscillatory motion. We evaluate retrieval on two real video data sets, and on a video data set exhibiting controlled variation in shape, color, motion and clutter.

show abstract

Section: Related Workmentioning

confidence: 99%

Reverse storyboarding for video retrieval

McNeill

Collomosse

2007

IET 4th European Conference on Visual Media Production (CVMP 2007)

View full text Add to dashboard Cite

show abstract

“…3 . First, statistics m, c , " are gathered within each tile and then a method similar to pyramid-linking" 11 is used to merge tiles into locales.…”

Section: Feature Localization Vs Image Segmentationmentioning

confidence: 99%

“…3 The idea is that a user would select a particular model to search for in a library of images or videos | I remember the instructor waved this book, but where is the video for that lecture? "…”

Section: Introductionmentioning

confidence: 99%

“…A coarse, e ective localization was obtained simply by accumulating pixels at the level of tiles into similar regions, based on color and eccentricity, called locales". 3 Here this idea can be extended to the growth of regions with similar chromaticity and compact eccentricity. Locales need not be mutually exclusive and thus do not form a true segmentation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

<title>Feature localization and search by object model under illumination change</title>

Drew

Tauber

1999

SPIE Proceedings

Self Cite

View full text Add to dashboard Cite

Color object recognition methods that are based on image retrieval algorithms can handle changes of illumination via image normalization, e.g. simple color-channel-normalization 1 or by forming a doubly-stochastic image matrix. 2However these methods fail if the object sought is surrounded by clutter. Rather than directly trying to nd the target, a viable approach i s t o g r o w a small number of feature regions called locales. 3 These are de ned as a nondisjoint coarse localization based on image tiles. In this paper, locales are grown based on chromaticity, which i s more insensitive to illumination c hange than is color. Using a diagonal model of illumination change, a least-squares optimization on chromaticity recovers the best set of diagonal coe cients for candidate assignments from model to test locales stored in a database. If locale centroids are also stored then, adapting a displacement model to include model locale weights, transformed pose and scale can be recovered. Tests on databases of real images show promising results for object query.

show abstract

“…All of them seem to either use global brightness normalization or simply do color quantization without any normalization, e.g., histogram binning. They include, Blobworld [5] that uses histogram binning to describe color features, C-bird [6] that uses normalization of each RGB channel before computing its chromaticity feature descriptor, ImageRETRO [7] that uses a 15 bin hue histogram as one of the color features, MARS [8] that uses a 2D histogram over the HS coordinates of the HSV space to represent color features, and two histograms one measuring coarseness other the directionality of the image to represent texture features, QBIC [10] uses a 256 dimensional RGB color histogram as one of the color features, Surfimage [11] uses RGB color histogram, edge orientation histogram as one of the low level features, WISE [12] in the preprocessing step rescales images and converts from RGB color space to another color space using normalization. We decided to compare our results with PicToSeek, because of its good performance, and demonstrate the effects of normalization.…”

Section: Introductionmentioning

confidence: 99%

<title>Image retrieval and reversible illumination normalization</title>

2005

View full text Add to dashboard Cite

We propose a novel approach to retrieve similar images from image databases that works in the presence of significant illumination variations. The most common method to compensate for illumination changes is to perform color normalization. The existing approaches to color normalization tend to destroy image content in that they map distinct color values to identical color values in the transformed color space. From the mathematical point of view, the normalization transformation is not reversible.In this paper we propose to use a reversible illumination normalization transformation. Thus, we are able to compensate for illumination changes without any reduction of content information. Since natural illumination changes affect different parts of images in different amounts, we apply our transformation locally to sub-images. Basic idea is to divide an image into sub-images, normalize each one separately, and then project it to an n-dimensional reduced space using principal component analysis. This process yields a normalized texture representation as a set of n-vectors. Finding similar images is now reduced to computing distances between sets of n-vectors. Results were compared with a leading image retrieval system.

show abstract

Illumination Invariance and Object Model in Content-Based Image and Video Retrieval

Cited by 44 publications

References 29 publications

Reverse storyboarding for video retrieval

Reverse storyboarding for video retrieval

<title>Feature localization and search by object model under illumination change</title>

<title>Image retrieval and reversible illumination normalization</title>

Contact Info

Product

Resources

About