One dimensional multi-resolution Local Binary Patterns features (1DMRLBP) for regular electrocardiogram (ECG) waveform detection

Abstract

Feeding a noisy signal to a biometric system degrades its performance. Hence, signal quality measure is used to avoid passing irregular signals to subsequent systems such as biometric systems. To tackle this issue, 1DMRLBP features, which are 1 dimensional signal feature extraction (inspired by the 2 dimensional image Local Binary Patterns) is proposed. 1DMRLBP with its multi-resolution capability captures local and global signal characteristics; and with its histogram extraction avoids segments misalignment and reduces the number of features. Also with some modifications, 1DMRLBP accommodates the problem of unknown amplitude of a signal. 1DMRLBP achieves 91% performance rate in distinguishing between regular and irregular ECG waveforms. MATLAB code and more information are available at www.comm.utoronto.ca/ ∼ wlouis/1DMRLBP.

Multimodal biometric system for ECG, ear and iris recognition based on local descriptors

“…Experiments demonstrated that the features extracted from one lead can be used to identify a person with 98% accuracy. To date, ECG has seen significant attention in biometric recognition studies [8,25,34,54], using both fiducial (specific anchor points such as P-QRST) and non-fiducial features. Despite the research efforts directed at ECG biometrics, at present there is only one commercial system available which exploits ECG or EMG biometrics for recognition, the B-secure [28].…”

“…Until now, one of the most popular strategies is fusion modalities. Thus, methods based on score fusion have achieved remarkable performance, most of which fuses two modalities [11,25,34]. As for ECG with the conventional biometric, there are fewer studies.…”

“…Researchers demonstrate the robustness of local texture descriptors in the multimodal biometric system especially in real-world applications as well as easy data acquisition [14,27,34]. The two major steps involved in biometric system are feature extraction and classification step.…”

“…to get features that will be useful in decision and identification of subjects. Although many studies on handcraft features have appeared in recent years, local binary pattern remains an interesting feature extraction and keep attracting the researchers particularly after his performance on biometric applications such as face, iris and ear [21,34,56]. In this purpose, three local texture descriptors namely 1D-LBP, Shifted 1D-LBP and 1D-MR-LBP have been implemented and detailed in this paper in an attempt to evaluate our multimodal biometric system.…”

“…Chatlani et al were the first who develop speech systems by using 1D-LBP to distinguish the important features from speech signals and proved its speed and applicability to a real problem which is computationally inexpensive [14]. Louis et al [34] demonstrate the use of 1D-MR-LBP to reduce noise and preserves the morphology of ECG waveforms. For all these multiple strengths and reasons, the use of these methods will drive our system to achieve the underlying targets.…”