Soowoong Jeong

The single image dehazing based on efficient transmission estimation

In this paper, we propose a novel method for estimating the transmission and removing the haze from single image. This method is based on observation that the property of haze is widely spread. Thus its estimated transmission should be smoothly changed over the scene. We employed the local entropy and log function for estimating the atmospheric light and the smooth transmission. Experimental results demonstrated that the proposed method can be applied efficiently to outdoor vision applications or devices including cameras and camcorders with low complexity.

Efficient naturalness restoration for non-uniform illumination images

This study presents a novel image-naturalness restoration for the purpose of achieving better image quality with respect to the human visual system. Specifically, the authors suppress the artefacts around edges using image gradient components. They also appropriately enhance the image brightness using an adaptive gamma correction. In particular, they simply create a mapping curve to adjust the contrast, and then perform contrast enhancement. Experimental results demonstrate that the proposed algorithm outperforms other state-of-the-art algorithms in terms of subjective and objective quality measures, and even computational speed. Therefore they believe that the proposed approach can be a useful tool to improve the visual quality of images under non-uniform illumination.

Effective Visual Tracking Using Multi-Block and Scale Space Based on Kernelized Correlation Filters

Accurate scale estimation and occlusion handling is a challenging problem in visual tracking. Recently, correlation filter-based trackers have shown impressive results in terms of accuracy, robustness, and speed. However, the model is not robust to scale variation and occlusion. In this paper, we address the problems associated with scale variation and occlusion by employing a scale space filter and multi-block scheme based on a kernelized correlation filter (KCF) tracker. Furthermore, we develop a more robust algorithm using an appearance update model that approximates the change of state of occlusion and deformation. In particular, an adaptive update scheme is presented to make each process robust. The experimental results demonstrate that the proposed method outperformed 29 state-of-the-art trackers on 100 challenging sequences. Specifically, the results obtained with the proposed scheme were improved by 8% and 18% compared to those of the KCF tracker for 49 occlusion and 64 scale variation sequences, respectively. Therefore, the proposed tracker can be a robust and useful tool for object tracking when occlusion and scale variation are involved.

Content awareness-based color image enhancement

This paper presents a novel color image enhancement method to improve the image brightness based on retinex theory. Specifically, the proposed method employs an adaptive gamma correction on the illumination components based on pixels. We appropriately perform the brightness enhancement. Additionally, we restore the detail and color using reflectance components. Experiments demonstrate that the proposed method outperforms existing algorithm in terms of subjective assessment and processing speed. Therefore, it can be applied effectively to outdoor vision applications or digital devices including mobile cameras and camcorders with low complexity.

Impulse noise reduction using distance weighted average filter

Switching median filter is known as one of effective algorithms for impulse noise reduction. In this paper, we present an improved switching median filter by considering weighted neighboring pixel locations. Specifically, the proposed method generates a flag map using boundary discriminative noise detection(BDND) detector. Next, we conduct a noise reduction by estimating the local noise density. When the local noise density is low, a corrupted pixel is replaced with the median value of uncorrupted neighboring pixels. In contrast, when the density is high, a noise searching window size increases until the predefined conditions are met. Then, a noise pixel is corrected by the weighted average of the uncorrupted values. Experiment results show that the proposed method outperforms the existing methods by about 0.5-3.7 dB on average.

A single image enhancement using inter-channel correlation

A single image enhancement algorithm using inter-color channel relationship is presented. It is based on the fact that the property of an infrared range image is similar to that of red channel in a visible range color image. Specifically, the proposed method analyzes the image details, which are mostly resided in dark area and not visible well in the visible range image, directly from the red channel of the given image. And they are used as a guidance to generate a weighting map for image enhancement. Experiment results show that the proposed scheme produces good outcomes in terms of revealing dark region details and visually pleasant observation.