Wen Chung Kao

Local contrast enhancement and adaptive feature extraction for illumination-invariant face recognition

Recognizing human faces in various lighting conditions is quite a difficult problem. The problem becomes more difficult when face images are taken in extremely high dynamic range scenes. Most of the automatic face recognition systems assume that images are taken under well-controlled illumination. The face segmentation as well as recognition becomes much simpler under such a constrained condition. However, illumination control is not feasible when a surveillance system is installed in any location at will. Without compensating for uneven illumination, it is impossible to get a satisfactory recognition rate. In this paper, we propose an integrated system that first compensates uneven illumination through local contrast enhancement. Then the enhanced images are fed into a robust face recognition system which adaptively selects the most important features among all candidate features and performs classification by support vector machines (SVMs). The dimension of feature space as well as the selected types of features is customized for each hyperplane. Three face image databases, namely Yale, Yale Group B, and Extended Yale Group B, are used to evaluate performance. The experimental result shows that the proposed recognition system give superior results compared to recently published literatures.

Image quality improvement for electrophoretic displays by combining contrast enhancement and halftoning techniques

The electrophoretic display (EPD) has dominated the panel used in electronic readers. However, getting more than 16 observable gray levels on an EPD is still a big challenge. A true color image should be converted into a 16 gray level image for displaying it on an EPD. In this paper, we propose an image processing tool that combines contrast enhancement and the modified Nasik pattern halftoning to deal with the problems in the image conversion chain. The experimental results show that the processed image has preserved more image textures after the image data are quantized into 16 gray levels1.

Configurable timing controller design for active matrix electrophoretic display

To design various electronic papers, the panel designers may adopt different types of electrophoretic materials and active matrix backplanes for resolution, speed, and contrast ratio considerations. The electrical and optical characteristics are quite different under different design requirements so that a good display timing controller should provide an easy way to modify waveforms accordingly. In this paper, we present a configurable timing controller for active matrix electrophoretic displays (EPD). The design principle of driving waveforms is first discussed and the controller can support all variations of driving waveforms with pulse width modulation. The performance of the controller has been verified by several practical EPD systems.

Design considerations of color image processing pipeline for digital cameras

Although many individual image processing steps have been well addressed in the field, very few good image pipeline designs were proposed to integrate these processing stages. In this paper, a new color image processing pipeline (IPP), which processes the image raw data captured from CCD/CMOS sensors and converts to the final color with exposure corrected, is presented, it bridges the gap between theoretical color imaging sciences and practical IPP implementation issues found in digital imaging systems. We demonstrate the IPP by processing different scenes of pictures and the results show that the proposed pipeline performs well in diverse scenes and illuminants

High Dynamic Range Imaging by Fusing Multiple Raw Images and Tone Reproduction

This paper presents an integrated color imaging system for taking images in extremely high dynamic range scenes. The system first fuses several differently exposed raw images to acquire more intensity information. The effective dynamic range of the image raw data can be extended to 256 times if five differently exposed images are fused. Then it runs edge detection iterations to extract the image details in different luminance levels. The proposed tone reproduction algorithm equalizes the histogram of the extracted fine edges which tends to assign larger dynamic range for highly populated regions. Finally, the local contrast enhancement is performed to further refine the image details. The experimental results show that the proposed high dynamic range imaging system has good performance on both tone and color reproduction.

Automatic phonocardiograph signal analysis for detecting heart valve disorders

Research highlights? This paper presents system for automatically determining the heart condition. ? The cycle segmentation is based on autocorrelation of heart beats. ? The feature extraction includes the STFT, DCT, and the adaptive feature selection. ? Only a few specific ones are selected for the classification of each hyperplane. Skilled cardiologists probe heart sounds by electronic stethoscope through human ears, but interpretations of heart sounds is a very special skill which is quite difficult to teach in a structured way. Because of this reason, automatic heart sound analysis in computer systems would be very helpful for medical staffs. This paper presents a complete heart sound analysis system covering from the segmentation of beat cycles to the final determination of heart conditions. The process of heart beat cycle segmentation includes autocorrelation for predicting the cycle time of a heart beat. The feature extraction pipeline includes stages of the short-time Fourier transform, the discrete cosine transform, and the adaptive feature selection. Many features are extracted, but only a few specific ones are selected for the classification of each hyperplane based on a systematic approach. The experiments are done by a public heart sound database released by Texas Heart Institute. A very promising recognition rate has been achieved.

Mltistage bilateral noise filtering and edge detection for color image enhancement

Removing noise while preserving and enhancing edges is one of the most fundamental operations of image/video processing. When taking pictures with digital cameras, it is frequently found that the color images are corrupted by miscellaneous noise and, hence, noise filtering is necessary. The difficulty is that usually the filtering will reduce the sharpness of the image. On the other hand, optical lens imperfections are usually equivalent to spatial low pass filters and tend to result in blurred images. It is customary to apply edge enhancement algorithm on the image in order to improve the sharpness, but this process usually increase the noise level as a by-product. In this paper, we present a new integrated approach to address these issues. The proposed approach combines color edge detection, bilateral noise filter, and edge enhancement based on suitable color spaces. The experimental results show that the proposed approach can effectively reduce the noise while preserving and enhancing edges.

Detection of cardiac arrhythmia in electrocardiograms using adaptive feature extraction and modified support vector machines

The electrocardiogram (ECG) analysis is one of the most important approaches to cardiac arrhythmia detection. Many algorithms have been proposed, however, the recognition rate is still unsatisfactory due to unreliable feature extraction in signal characteristic analysis or poor generalization capability of the classifier. In this paper, we propose a system for cardiac arrhythmia detection in ECGs with adaptive feature selection and modified support vector machines (SVMs). Wavelet transform-based coefficients and signal amplitude/interval parameters are first enumerated as candidates, but only a few specific ones are adaptively selected for the classification of each class pair. A new classifier, which integrates k-means clustering, one-against-one SVMs, and a modified majority voting mechanism, is proposed to further improve the recognition rate for extremely similar classes. The experimental results show that the proposed ECG analysis approach can obtain a higher recognition rate than the published approaches. By testing the system with more than 100,000 samples in MIT-BIH arrhythmia database, the average recognition rate is 98.92%, and the recognition rate for each class is kept above 92%.

Configurable timing controller design for active matrix electrophoretic display with 16 gray levels

To design various electronic papers, the panel designers may adopt different types of electrophoretic materials and active matrix backplanes for resolution, speed, and contrast ratio considerations. The electrical and optical characteristics are different for various application conditions so that a good timing controller should provide an easy way to modify waveforms accordingly. In this paper, we present a configurable timing controller for active matrix electrophoretic displays. It can support all variations of driving waveforms that are based on pulse width modification. The performance of the controller has been applied on several practical EPD systems whose resolutions are not identical.

Electrophoretic Display Controller Integrated With Real-Time Halftoning and Partial Region Update

Electrophoretic displays (EPDs) have become the main solution to electronic paper because of its invariably reflective and bistable properties. With different requirements of display resolution, speed, and contrast ratio, the electronic paper (E-paper) designers may adopt different types of electrophoretic material and active matrix backplane for building the system. The driving waveforms should be also customized for accommodating different application specifications and environment temperature. In this paper, we present a configurable EPD controller which can support various application specifications. In addition to providing a new architecture of compacted waveform lookup tables, the controller supports partial image update and integrates a real-time halftoning engine for improving image quality. The design has been implemented on a field-programmable gate array (FPGA) and tested with several real EPDs.