Jinho Park

Lens distortion correction and enhancement based on local self-similarity for high-quality consumer imaging systems

In this paper, a novel image enhancement system for a wide-angle lens camera is presented. The proposed system consists of; i) lens distortion correction using space-varying interpolation kernels and ii) image restoration based on the local self-similarity. The correction process for the geometric distortion produced by a wide-angle lens results in radial distortion artifacts caused by non-linear resampling. To reduce such artifacts, the proposed algorithm uses space-varying interpolation kernels derived from the lens calibration data. The corrected image is further enhanced using self-example-based image restoration. Experimental results demonstrate the proposed method can correctly remove the geometric distortion and further enhance the quality of the radially interpolated image.

Extended fisheye lens model for practical geometric correction and image enhancement

An extended fisheye lens model is presented to control the size ratio between the distorted and virtually undistorted images based on orthographic projection. The optimum size ratio is derived to correct the barrel distortion of a fisheye lens so that the maximum amount of the peripheral region is reconstructed with the minimum visual distortion. The geometric correction generates an aliasing artifact in the central region and a jagging artifact in the peripheral region. Based on the proposed lens model, a novel image enhancement algorithm is also presented to remove the aliasing and jagging artifacts in the geometrically corrected image. Experimental results demonstrate that the proposed enhancement method outperforms existing methods in the sense of objective and subjective measures.

Automatic top-view transformation for vehicle backup rear-view camera

An automatic top-view transformation method is presented for a vehicle backup rear-view camera. The proposed method consists of two steps: i) automatic corresponding points estimation based on the lens specification and ii) view transformation based on the direct linear transform algorithm. Major contribution of this work is automatic view transformation that is optimized for the vehicle rear view camera system. The proposed method can be applied to various imaging systems, such as automotive imaging systems, intelligent surveillance systems, and vehicle rear view cameras.

Three-dimensional object motion and velocity estimation using a single computational RGB-D camera.

In this paper, a three-dimensional (3D) object moving direction and velocity estimation method is presented using a dual off-axis color-filtered aperture (DCA)-based computational camera. Conventional object tracking methods provided only two-dimensional (2D) states of an object in the image for the target representation. The proposed method estimates depth information in the object region from a single DCA camera that transforms 2D spatial information into 3D model parameters of the object. We also present a calibration method of the DCA camera to estimate the entire set of camera parameters for a practical implementation. Experimental results show that the proposed DCA-based color and depth (RGB-D) camera can calculate the 3D object moving direction and velocity of a randomly moving object in a single-camera framework.

Non-dyadic lens distortion correction and enhancement of fish-eye lens images

In this paper, a novel image enhancement method is presented for a fish-eye lens camera. In order to remove jagging and blur artifacts that are generated by a correction process of geometric lens distortion, the proposed method searches similar patches in a non-dyadic scale space. For further enhancing the artifact-removed image, a self-example-based image restoration is also used. Experimental results show that the proposed method can successfully remove not only geometric distortion of a fish-eye lens but also jagging and blur artifacts in the corrected image.

Hybrid Sensor Network-Based Indoor Surveillance System for Intrusion Detection

This paper presents a novel hybrid sensor-based intrusion detection system for low-power surveillance in an empty, sealed indoor space with or without illumination. The proposed system includes three functional steps: (i) initial detection of an intrusion event using a sound field sensor; (ii) automatic lighting control based on the detected event, and (iii) detection and tracking the intruder using an image sensor. The proposed hybrid sensor-based surveillance system uses a sound field sensor to detect an abnormal event in a very low-light or completely dark environment for 24 h a day to reduce the power consumption. After detecting the intrusion by the sound sensor, a collaborative image sensor takes over an accurate detection and tracking tasks. The proposed hybrid system can be applied to various surveillance environments such as an office room after work, empty automobile, safety room in a bank, and armory room. This paper deals with fusion of computer-aided pattern recognition and physics-based sound field analysis that reflects the symmetric aspect of computer vision and physical analysis

Iterative refinement of transmission map for stereo image defogging

This paper presents a novel stereo image defogging algorithm using the disparity map. The proposed algorithm iteratively performs three steps; i) disparity map estimation using the optical flow, ii) transmission map generation using the estimated disparity map, and iii) reconstruction of defogged image using the transmission map. Experimental results show that the proposed method can successfully remove the fog without color distortion or ringing artifacts.

Iterative Refinement of Transmission Map for Stereo Image Defogging Using a Dual Camera Sensor

Recently, the stereo imaging-based image enhancement approach has attracted increasing attention in the field of video analysis. This paper presents a dual camera-based stereo image defogging algorithm. Optical flow is first estimated from the stereo foggy image pair, and the initial disparity map is generated from the estimated optical flow. Next, an initial transmission map is generated using the initial disparity map. Atmospheric light is then estimated using the color line theory. The defogged result is finally reconstructed using the estimated transmission map and atmospheric light. The proposed method can refine the transmission map iteratively. Experimental results show that the proposed method can successfully remove fog without color distortion. The proposed method can be used as a pre-processing step for an outdoor video analysis system and a high-end smartphone with a dual camera system.

Improved DCP-based image defogging using stereo images

Image defogging has recently received attentions in many applications such as advanced drive assistance systems (ADAS) and intelligent surveillance systems to acquire a high-quality images. This paper presents a novel depth-based image defogging method using stereo images. The depth information obtained from a pair of stereo foggy images and then fuzzy C-mean (FCM) clustering is applied to reduce matching errors caused by atmospheric absorption and scattering during light propagation. The estimated depth information is used as weighting values in the dark channel prior (DCP)-in the defogging process. Experimental results show that the proposed method can successfully remove foggy components in the image without color distortion.

Motion Estimation-based Human Fall Detection for Visual Surveillance

Currently, the world’s elderly population continues to grow at a dramatic rate. As the number of senior citizens increases, detection of someone falling has attracted increasing attention for visual surveillance systems. This paper presents a novel fall-detection algorithm using motion estimation and an integrated spatiotemporal energy map of the object region. The proposed method first extracts a human region using a background subtraction method. Next, we applied an optical flow algorithm to estimate motion vectors, and an energy map is generated by accumulating the detected human region for a certain period of time. We can then detect a fall using k-nearest neighbor (kNN) classification with the previously estimated motion information and energy map. The experimental results show that the proposed algorithm can effectively detect someone falling in any direction, including at an angle parallel to the camera’s optical axis.