Richard M. Dansereau

Carleton University, Systems and Computer Engineering, Faculty Member

Followers

Following

Co-authors

Public Views

Interests

uploads

Papers

Robust Joint Audio-Video Localization in Video Conferencing Using Reliability Information

IEEE Transactions on Instrumentation and Measurement, Aug 1, 2004

Download

Time Domain Method for Precise Estimation of Sinusoidal Model Parameters of Co-Channel Speech

Research Letters in Signal Processing, 2008

Download

An Efficient Measure for Nonlinear Distortion Severity due to HPA in Downlink DS-CDMA Signals

Eurasip Journal on Wireless Communications and Networking, Oct 10, 2010

Download

Hybrid spread spectrum orthogonal waveforms for MIMO radar

In multiple input multiple output (MIMO) radar systems, choosing a proper orthogonal waveform is ... more In multiple input multiple output (MIMO) radar systems, choosing a proper orthogonal waveform is a critical task. A new hybrid spread spectrum (HSS) technique is proposed to maintain orthogonality at the transmit and receive ends. The HSS technique is a combination of direct-sequence spreading and frequency hopping schemes. In the context of MIMO radar, the transmitted signals are first spread using Hadamard-Walsh orthogonal codes and in every pulse repetition period, each signal hops to a different center frequency. The transmitted HSS signals are orthogonal in frequency and code domains. Simulation results show that the proposed HSS technique can achieve sharper auto ambiguity response and lower sidelobe cross ambiguity response with a gain of over 10 dB and better probability of detection in comparison with frequency orthogonality technique. The proposed HSS technique has the potential to resolve closely spaced targets and provide better immunity against narrowband interferences.

Improved Covariance Matrix Estimation using Riemannian Geometry for Beamforming Applications

The estimation of interference plus noise covariance (INC) matrix for beamforming applications is... more The estimation of interference plus noise covariance (INC) matrix for beamforming applications is considered from a Riemannian space perspective. A new INC estimation technique based on regularized Burg algorithm (RBA), Riemannian mean and Riemannian distance is proposed to maintain a stable performance in presence of angle of arrival mismatch and small sample size with high and low signal to interference plus noise ratio (SINR). The RBA is exploited to generate Toeplitz Hermitian positive definite (THPD) covariance matrices from the estimates of the reflection coefficients for each radar snapshot. The estimated INC is formulated as a linear combination of THPD covariance matrices of the interference plus noise excluding potential target snapshots. The weights of the linear combination operation are based on the Riemannian distance between the Riemannian mean and each THPD covariance matrix. The largest distance (potential target) will have zero weight and the smallest distance will have maximum weight. Simulation results demonstrate the performance of the proposed technique in comparison with sample covariance and Riemannian mean covariance under steering vector mismatch and small sample size in presence of high and low SINR.

Random walks for image segmentation containing translucent overlapped objects

Image segmentation methods usually separate an image into non-overlapping regions. However, in so... more Image segmentation methods usually separate an image into non-overlapping regions. However, in some data translucent regions partially overlap. Having overlapped regions will cause methods not designed for overlapped objects to perform poorly or not work at all. The random walker (RW) image segmentation algorithm shows interesting results and is widely used in many applications. It creates a graph from the image and segments the non-overlapped regions based on the similarity of nodes in the graph while keeping the results smooth over neighboring nodes. In this paper, we propose a method based on RW that can segment translucent overlapped regions. We create a multi-layer graph from a 2D translucent overlapped input image and generate the Laplacian matrix based on this multilayer graph. Having this new Laplacian matrix and providing a small number of user-defined labels (seeds), the algorithm segments the input image into regions. Experimental results on synthetic images show the strength of our proposed method with more than 98% segmentation accuracy.

STP-DEQ-Net: A Deep Equilibrium Model Based on ISTA Method for Image Compressive Sensing

2022 30th European Signal Processing Conference (EUSIPCO), Aug 29, 2022

Compressive Sensing using S-transform in Pulse Radar

Compressive sensing is a novel approach for data acquisition at a sub-Nyquist rate. In this paper... more Compressive sensing is a novel approach for data acquisition at a sub-Nyquist rate. In this paper, we propose an algorithm using S-transform basis functions to adaptively sample pulse radar signals based on compressive sensing. Based on prior information from previously received pulse intervals, the algorithm adaptively samples the current interval at different resolutions. Exploiting the S-transform's ability to extract phase information and its time and frequency localization properties, the proposed algorithm achieves target range and Doppler frequency detection in a significantly reduced processing time compared to competitor approaches. The simulation results using the receiver operating characteristic (ROC) curves shows the proposed algorithm performance in different scenarios.

Deep Learning Methods for Image Segmentation Containing Translucent Overlapped Objects

Download

Learning-Based multilabel random walks for image segmentation containing translucent overlapped objects

2017 40th International Conference on Telecommunications and Signal Processing (TSP), 2017

Supervised image segmentation methods usually start with information extracted from the learning ... more Supervised image segmentation methods usually start with information extracted from the learning phase to separate an image into non-overlapping regions. We have used user input information or seeds in our previous work to segment partially overlapped translucent regions. However providing a lot of seeds might sometimes be too time consuming that might make the method perform poorly or not work at all. Machine learning algorithms consist of two major phases: learning phase where the information will be generated based on the data, and test phase where the generated information will be used to improve the performance of the method. In previous work user guided labels were used as hard seeds in the RW algorithm. In this paper we extend our previous work to be able to segment multilabel translucent overlapped objects using soft seed information. We first map each segment as a class on a 25D manifold in the learning phase. Then the probability of assigning each of the image pixels to the segments, data term, is obtained by calculating the geodesic distance between the pixels' features and these classes on the manifold. This data term is then used as soft seeds in the RW algorithm instead of user predefined labels. Experimental results on synthetic images show the strength of our proposed method comparing to our previous algorithm with more than 95% segmentation accuracy.

Fast Reconstruction of 1D Compressive Sensing Data Using a Deep Neural Network

International Journal of Signal Processing Systems, 2020

Download

Image Registration Under Illumination Variations Using Region-Based Confidence Weighted $M$-Estimators

IEEE transactions on image processing, Mar 1, 2012

Download

$Research paper thumbnail of Progressive transmission of images through fractal surface interpolation$

Progressive transmission of images through fractal surface interpolation

ABSTRACT

$Research paper thumbnail of New relative multifractal dimension measures$

New relative multifractal dimension measures

$Research paper thumbnail of Perceptual image compression through fractal surface interpolation$

Perceptual image compression through fractal surface interpolation

This paper presents a perceptual image compression technique using fractal surface interpolation ... more This paper presents a perceptual image compression technique using fractal surface interpolation (FSI). This technique relies on the extraction and reconstruction of self-affine fractal surfaces with a specified Hurst exponent H*. Fractional Brownian motion (fBm) through midpoint displacement (MPD) provides the basis for generating these fractal surfaces between interpolation points. The MPD algorithm is designed to eliminate creasing often associated with MPD in two dimensions. Experimental results show that when applied to 512×512 8-bit gray-level images, perceptually good results are achieved with a bit rate of 0.35 bpp