Home Home
M. Ulises Ramos Sánchez, University of Surrey (U.K.)
Jiri Matas, University of Surrey (U.K.)
Josef Kittler, University of Surrey (U.K.)
We present a statistical, colour-based technique for lip tracking intended to support personal verification. The lips are automatically localised in the original image by exploiting grey-level gradient projections as well as chromaticity models to find the mouth area in an automatically segmented region corresponding to the face area. A B-spline, initially with an elliptic shape is then generated to start up tracking. Tracking proceeds by estimating new lip contour positions according to a statistical chromaticity model for the lips. These measurements are used together with a Lagrangian formulation of contour dynamics to update the new spline control points. The method has been tested on the M2VTS database, where lips were accurately tracked on sequences of speaking subjects consisting of more than hundred frames. The tracker can be used to perform feature extraction from the mouth area as well as for model detection for personal verification applications.
ic972973.pdf ic972973.pdf
TOP
Man-Nang Chong, Nanyang Technological University (Singapore)
P. Liu, Nanyang Technological University (Singapore)
W.P. Goh, Nanyang Technological University (Singapore)
Dilip Krishnan, Nanyang Technological University (Singapore)
This paper proposes a new spatial-temporal MRF model for the detection of missing data (also referred to as blotches) in image sequences. The blotches in noise-corrupted image sequences exhibit temporal discontinuity characteristic which is commonly used for the detection of blotches. However, the badly motion compensated pixels will also appear as temporal discontinuities, thus making it difficult to distinguish the true blotches from the poorly motion compensated regions. The proposed MRF model addresses the problem of incorrect detection. It is found that the degree of false-alarm in the detection of the blotches in image sequences can be reduced by using a moving-edge detector in the MRF model to identify the blotch-edges from the moving-edges.
ic972977.pdf TOP
Luis Torres, UPC (Spain)
David García, UPC (Spain)
Anna Mates, UPC (Spain)
The objective of this paper is to provide a robust representation of moving images based on layers. To that goal, efficient motion estimation and segmentation techniques by affine model fitting suitable for the construction of layers have been designed. Layered representations are important in several applications. In particular, they are very appropriate for object tracking, object manipulation and content-based scalability which are among the main functionalities of the future standard MPEG-4. In addition a variety of examples are provided that give insight into the performance of the representation of moving images using layers.
ic972981.pdf ic972981.pdf TOP
Bo Tao, Princeton University (U.S.A.)
Bradley Dickinson, Princeton University (U.S.A.)
An adaptive watermarking technique is introduced in this work. A regional perceptual classifier is employed to assign a noise sensitivity index to each region. The watermark is inserted in the original image according to this index by using block DCT. The detection of the watermark is designed to achieve a desired false alarm probability.
ic972985.pdf ic972985.pdf TOP
Susie J. Wee, HP Labs (U.S.A.)
Michael O. Polley, TI (U.S.A.)
This paper investigates the problem of communicating video over a broadcast channel. The broadcast channel is expressed in terms of the channel capacity that exists between the transmitter and each receiver in the broadcast area -- Shannon's separation theorem does not apply for video communication over this class of channels. Digital (discrete time, discrete amplitude) and hybrid (discrete time, discrete/continuous amplitude) transmission and video coding methods are discussed. Joint source and channel coding principles are employed to effectively couple these methods to form efficient systems for communicating video over broadcast channels. A framework is presented for characterizing and bounding the performance of these systems; the results lead to interesting directions for future work.
ic972989.pdf ic972989.pdf TOP
Cesar A. Gonzales, IBM T.J. Watson Research Center (U.S.A.)
Ali N. Akansu, New Jersey Institute of Technology, ECE Dept. New Jersey Center for Multimedia Research (U.S.A.)
Subband coding has been successfully used for image and video applications. This paper emphasizes two design issues of a subband codec. Namely, the selection of the filterbank family and the width of zero-zones in zero-zone linear quantizers. It is shown that the shift only filters (Multiplierless PR-QMF) performs as good as other popular filterbank families. Finally, the significance of the quantizers' zero-zones in codec's rate-distortion performance is presented. It is found that the proposed efficient subband image/video codec performs comparable with the leading techniques like embedded zero-tree wavelet (EZW) coding scheme reported in the literature. We also incorporated this efficient subband codec into a motion compensated video coding algorithm. It is observed that the superior performance of zero-zone linear quantizers is also valid for subband video coding. The subjective performance of the proposed efficient codec in image and video coding are found consistent with the objective performance results obtained.
ic972993.pdf ic972993.pdf TOP
Austin Y. Lan, University of Washington (U.S.A.)
Jenq-Neng Hwang, University of Washington (U.S.A.)
The MPEG video compression standard effectively exploits spatial, temporal, and coding redundancies in the algorithm. In its generic form, however, only a minimal amount of scene adaptation is performed. Video can be further compressed by taking advantage of scenes where the temporal statistics allow larger inter-reference frame distances. This paper proposes the use of motion analysis (MA) to adapt to scene content. The actual picture type (I, P, or B) decision is made by examining the accumulation of motion measurements since the last reference frame was labeled. Depending on the video content, this proposed algorithm can achieve from 2% to 13.9% savings in bits while maintaining similar quality.
ic972997.pdf TOP
Christian Derviaux, University of Valenciennes (France)
Francois Xavier Coudoux, University of Valenciennes (France)
Marc Gazalet, University of Valenciennes (France)
Patrick Corlay, University of Valenciennes (France)
Mohamed Gharbi, University of Valenciennes (France)
One of the drawbacks of the Discrete Cosine Transform (DCT) is visible block boundaries due to coarse quantization of the coefficients. In this paper, an algorithm for the reduction of blocking artifacts is presented. The proposed method allows to produce higher quality reconstructed images by adaptively filtering the video signal according to the noise visibility. A visual model is therefore defined for predicting the block edge visibility across each picture of the coded sequence. This model that accounts for the perceptual characteristics of the block distortion is widely described in the main section of the paper. Then experimental results are presented for low bit-rate coded sequences ; they show that the postfiltering operation yields significant results with enhanced visual quality.
ic973001.pdf ic973001.pdf TOP
Pekin Erhan Eren, University of Rochester (U.S.A.)
Yucel Altunbasak, University of Rochester (U.S.A.)
Ahmet Murat Tekalp, University of Rochester (U.S.A.)
This paper presents region-based affine motion parameter clustering methods by motion-vector and intensity matching to provide improved robustness and alignment of motion boundaries with real object boundaries. Regions may be formed as fixed- or variable-size rectangular blocks, triangular patches, or arbitrary-shaped areas defined by color or texture uniformity. A particular combination of these affine clustering methods is then proposed to obtain the best segmentation results on a variety of image sequences. Experimental results showing the benefits of the proposed scheme are provided.
ic973005.pdf ic973005.pdf TOPMehdef Eyvazkhani, Ecole Nat. Sup. des Telecomm. (France)
To transmit a video signal over the broadband networks, the first step is to compress the video stream in order to reduce the bit rate. However, the lossy nature of the broadband networks will degrade the quality of loss sensitive applications as compressed video signal. In this paper, a video subband coding system is discussed, which allows us to compensate the information losses due to packet loss. This subband coding system treats the video over its three dimensions. one in time and two others in spatial dimensions. For the two spatial dimensions, the non-separable filter bank is used which provides the better performances. A main principle idea in this coding system is that several channels with different priorities are offered. This novel scheme propose to transmit the video signal over the network with different levels of protection. Thus, even the packet loss occurs in the network, the most important part of information will be transmitted over a very protected channel and the rest of information whic
ic973009.pdf TOP
Jean-Pierre Leduc, Georgia Institute of Technology (U.S.A.)
Fernando Mujica, Georgia Institute of Technology (U.S.A.)
Romain Murenzi, Clark Atlanta University (U.S.A.)
Mark J.T. Smith, Georgia Institute of Technology (U.S.A.)
This paper addresses the problem of detecting and tracking moving objects in digital image sequences. The main goal is to detect and select mobile objects in a scene, construct the trajectories, and eventually reconstruct the target objects or their signatures. It is assumed that the image sequences are acquired from imaging sensors. The method is based on spatio-temporal continuous wavelet transforms, discretized for digital signal analysis. It turns out that the wavelet transform can be used efficiently in a Kalman filtering framework to perform detection and tracking. Several families of wavelets are considered for motion analysis according to the specific spatio-temporal transformation. Their construction is based on mechanical parameters describing uniform motion, translation, rotation, acceleration, and deformation. The main idea is that each kind of motion generates a specific signal transformation, which is analyzed by a suitable family of continuous wavelets. The analysis is therefore associated with a set of operators that describe the signal transformations at hand. These operators are then associated with a set of selectivity criteria. This leads to a set of filters that are tuned to the moving objects of interest.
ic973013.pdf ic973013.pdf TOP
Riccardo Leonardi, SCL-DEA/UNIBS (Italy)
Andrea Mazzarri, SCL-DEA/UNIBS (Italy)
Alberto Signoroni, SCL-DEA/UNIBS (Italy)
In this work a new way to improve the representation of images using a DWT for coding purposes is presented. The idea lies in combining all wavelet coefficients related to detail information at a same resolution level but along the different orientations, into a single image. Given that detail information is located for all subband in the neighborhood of high frequency textures or edge locations, the pattern of significant coefficients remains unchanged after the combination process. This process allows to reduce the number of transformed coefficients, while preserving the multiresolution structure. This information can thus be efficiently coded using a multiresolution embedded coding scheme. Overall, a higher coding efficiency can be reached while preserving the cross-scale prediction of significance among coefficients. Ultimately, approximate detail information must be recovered from the combined and coded data for each subband of the original wavelet, so as to reconstruct a decoded image.
ic973017.pdf ic973017.pdf TOP
Juan Domingo, Univ. de Valencia (Spain)
Guillermo Ayala, Univ. de Valencia (Spain)
María Elena Díaz, Univ. de Valencia (Spain)
This communication describes a method to detect and characterise the independent movement of several rigid objects in a sequence of images. No model is assumed for the moving objects, meanwhile a linear model in the image plane is accepted for the pixels of each object, which can be changed without loosing of generality. The approach works on each pair of images. Relevant points are detected on them and then each moving object is identified as two clusters of similar relevant points, one per image, that perform a consistent movement. Differently from previous approaches, the consistent movement constraint is incorporated into the cluster building phase. The main property of the proposed method is its robustness, since minimal assumptions are assumed. Results are provided with tests on two different kinds of scenes: a laboratory one and a natural traffic scene.
ic973021.pdf ic973021.pdf TOPAdnan M. Alattar, KFUPM (Saudi Arabia)
The use of fade in video production to smooth scene changes and enhance the video quality complicates subsequent video compression or video editing. It is important to detect the fade regions in order to improve the quality of the compressed video or to allow automatic parsing of the video for the purpose of editing and data-base indexing. In this paper a fade detector that exploits the changes in the average luminosities as well as the semi-parabolic behavior of the variances of the frames in the fade region is developed. Simulation results indicate that the developed detector has over a 95% reliability rate.
ic973025.pdf TOP