Joo-hyun Kim

Image quality enhancement by real-time gamma correction for a CMOS image sensor in mobile phones

Real-time gamma correction is an essential function in display devices such as CRTs, plasma TVs, and TFT LCDs. In this paper, we present the image quality enhancement by real-time gamma correction for a CMOS image sensor in mobile phones. The proposed algorithm and system operate with a block structure and the bit width of input and output of the proposed system is 12-bit. The proposed gamma system is reduced the error range and enhanced the image quality compared with conventional system that has 10-bit input and 8-bit output. The proposed system was implemented experimentally by using Xilinx Virtex4 FPGA and was successfully demonstrated with a CMOS image sensor for mobile application.

Auto white balance system using adaptive color samples for mobile devices

This paper presents a novel auto white balance (AWB) algorithm for images based on an YCbCr color space. The proposed algorithm divides all the pixels into fourteen groups by Y luminance and CbCr color components. For accuracy, we employ an adaptive color samples that can obtain the sum of the accumulated pixels up to the standard group for estimating a degree of white objectpsilas color changes. The color samples are used for the adjustment of the relative amounts of red, green, and blue primary colors. In order to avoid a field-flicker noise of video streams, the proposed algorithm adopts infinite impulse response (IIR) filters that vary the operating speed of overall AWB system. The performance of the algorithm is also verified by implementing the hardware of the system.

Real-time edge adaptive color interpolation for an ultra small HD-grade video sensor in mobile devices

To render a full-color image using a single image sensor, a color interpolation process is required to estimate the two missing color values at each pixel. In this paper, we present a real-time edge adaptive color interpolation method and hardware implementation for camera in mobile devices. The proposed method first calculates the edge indicators which consider horizontal, vertical and diagonal directions. At the same time, it modifies the inputted threshold values compares them with the edge indicators. Then, the proposed method decides the edge region and selects filter coefficients. Finally, the missing two color values are interpolated by using filter processing. The proposed system was implemented experimentally by using Xilinx Virtex4 FPGA and was successfully demonstrated with a CMOS sensor capturing a 720P (1280pixelstimes720lines pixels, 30fps) HD-grade video movie and a still image simultaneously.