Jong-Hyub Lee

An Illumination-Adaptive Colorimetric Measurement Using Color Image Sensor

An image sensor for a use of colorimeter is characterized based on the CIE standard colorimetric observer. We use the method of least squares to derive a colorimetric characterization matrix between RGB output signals and CIE XYZ tristimulus values. This paper proposes an adaptive measuring method to obtain the chromaticity of colored scenes and illumination through a 3×3 camera transfer matrix under a certain illuminant. Camera RGB outputs, sensor status values, and photoelectric characteristic are used to obtain the chromaticity. Experimental results show that the proposed method is valid in the measuring performance.

A novel stereoscopic image processing pipeline

This paper presents a novel stereoscopic image processing pipeline which ensures the consistency of a stereoscopic image pair. Although a stereoscopic image calibration method should resolve both photometric and geometric binocular asymmetries, most previous approaches [1], [2] have mainly focused on the geometric calibration so far. In contrast to this biased research, we propose both photometric and geometric calibration methods which are readily applicable to an embedded pipeline for a stereoscopic camera system. The experimental results demonstrate the performance of the proposed pipeline for calibrating a stereoscopic image pair.

A novel anti-vignetting method for color shading artifact suppression

In this paper, we present a novel automatic anti-vignetting method which effectively alleviates color shading artifacts regardless of ambient color temperature. The proposed method incorporates both a Color Temperature Metric (CTM) and a Vignetting Gain Control (VGC) algorithm in a commercial Image Signal Processor (ISP) which is embedded in mobile phone camera modules. Experimental results validate that our proposed method explicitly addresses the color shading problem in a wide variety of color temperatures.