Earticle

In this paper, we present an effective single image fog removal by using the Joint Bright and Dark Channel (JBDC) and pixel-based transmission estimation to enhance the visibility of outdoor images susceptible to degradation due to weather and environmental conditions. The conventional methods include refinement process of coarse transmission with heavy computational complexity. The proposed transmission estimation reveals excellent edge-preserving performance and does not require the refinement process. We estimate the atmospheric light in pixel-based fashion, which can improve the transmission estimation performance and visual quality of the restored image. Moreover, we propose an adaptive transmission estimation to enhance the visual quality specifically in sky regions. Comprehensive experiments on various fog images show that the proposed method exhibits reduced computational complexity and excellent fog removal performance, compared with the existing methods; thus, it can be applied to various fields including real-time devices.

We are intended to propose a video compression based on Modified Pixel-Based Motion Estimation (MPME) technique to solve the current problems in the literary works. The proposed system will be comprised of four phases, Displacement Calculation, Motion Estimation, Encoding and Decoding. In displacement calculation the I-frame and the P-frame has to been constructed from the video frames. In second phase, the motion vector has been finding out by using the Modified Pixel Based Motion estimation. In here we are creating four movement vector predictors (West Predictor, Neighbor Predictor, Median Predictor and Centre Predictor) with the help of diamond based search pattern. The proposed motion estimation technique is conducted on a pixel-by-pixel basis and the predicted P-frame has been developed using motion estimation technique. The residual is calculated with the original P-frame and the predicted P frame, and the residual is passed into next phase. After the motion vector estimation each frame is encoded by using JPEG-LS encoder. Next the encoded values of compressed video frames are decoded by using the decoding technique and for that JPEG-LS decoder is used. The proposed video compression using MPME will be implemented in the platform MATLAB and the performance of the video codec procedure is assessed by contrasting the consequence of proposed system with the conventional motion estimation algorithms.

In H.264, sub-pel motion estimation (ME) has strong effect when coding video sequences. 1/4-pel performs better at low bitrate while 1/8-pel gives better results at high bitrate. In this paper, a variable sub-pixel motion vector resolution based on block mode motion estimation method is proposed. Experiment results show that the proposed method can take the advantage of 1/4-pel at low bitrate and 1/8-pel at high bitrate. In addition to that, time is reduced from 14% to 53% compared to KTA1.3 with 1/8-pel motion vector (MV) resolution.

본 논문은 다시점 카메라로부터 획득된 다시점 영상을 이용하여, 영상내의 모든 화소에 대한 정확한 변이 정보를 획득하는 알고리듬을 제안한다. 제안한 방법은 정합창(measurement window)을 사용하는 화소기반 정합방법을 통해 정확한 변이 추정과 비용 함수 계산시 반복적인 계산을 줄임으로써 전체적인 수행 속도를 향상시킬 수 있다. 정확한 변이 정보를 획득하는데 있어서 가려진 영역이 많은 영향을 미친다. 또한 가려진 영역의 판단 및 처리는 2대의 카메라로는 한계가 있다. 따라서 본 논문에서는 평행식으로 구성된 5대 이상의 다시점 카메라로부터 획득된 다시점 영상을 사용함으로써 변이 정보의 신뢰도를 높이고, 가려진 영역을 처리할 수 있다. 또한 비용 함수 계산의 반복성을 이용하여 계산량을 줄임으로써 전체적인 수행시간을 줄인다. 이렇게 획득된 변이 정보는 다시점 영상의 중간시점영상 생성을 위해 사용된다.

본 논문에서는 안개 성분을 포함한 실외영상의 효과적인 응용을 위하여 저 복잡도를 갖는 전달량 추정 및 픽셀 기반 JBDCP(: Joint Bright and Dark Channel Prior)를 이용한 단일 영상 기반 안개 제거 방법을 제안한다. 기존의 전달량 추정은 계산량과 메모리 요구량이 큰 정련과정을 포함하는 반면, 제안하는 안개 제거 방법은 픽셀단위 및 블록단위 dark channel 정보를 결합하여 에지 정보가 보존되는 전달량을 추정하고 복잡도를 크게 감소시킨다. 또한 픽셀 기반 JBDCP를 이용하여 영상의 각 위치마다 다른 안개값을 구함으로써 영상의 특성을 반영한 전달량 추정이 가능하다. 다양한 안개 영상에 대해 수행한 실험 결과는 제안하는 방법이 기존의 방법에 비해 저 복잡도로 실행되면서 우수한 안개 제거 성능을 보여 실시간 기기를 포함한 다양한 분야에 응용될 수 있음을 나타낸다.

This paper proposes a single image dehazing that utilizes the transmission estimation with low complexity and the pixel-based JBDCP (Joint Bright and Dark Channel Prior) for the effective application of hazy outdoor images. The conventional transmission estimation includes the refinement process with high computational complexity and memory requirements. We propose the transmission estimation using combination of pixel- and block-based dark channel information and it significantly reduces the complexity while preserving the edge information accurately. Moreover, it is possible to estimate the transmission reflecting the image characteristics, by obtaining a different air-light for each pixel position of the image using the pixel-based JBDCP. Experimental results on various hazy images illustrate that the proposed method exhibits excellent dehazing performance with low complexity compared to the conventional methods; thus, it can be applied in various fields including real-time devices.

본 논문에서는 기상 및 환경조건에 영향을 받아 열화되기 쉬운 실외영상의 시인성을 개선하고 다양한 기기에 적용하기 위하여 저 복잡도의 단일 영상 안개 제거방법을 제안한다. 기존 방법에서는 거친 형태의 전달량을 추정한 후 연산량 및 메모리 요구량이 큰 정련 과정을 포함하는 반면, 제안하는 전달량 추정 방법은 에지 근처에서 픽셀 단위 dark channel과 패치 단위 dark channel을 비교함으로써 에지를 보존하는 특성이 우수하고 정련 과정이 필요하지 않아 저복잡도 전달량 추정이 가능하다. 또한, 픽셀 단위 JBDC(Joint Bright and Dark Channel)를 이용하여 각 픽셀마다 안개값을 예측함으로써 정밀한 전달량 추정과 영상의 특성에 따라 적응적인 안개 제거가 가능하다. 다양한 안개 영상에 대해 수행한 실험 결과는 제안하는 방법이 기존의 방법에 비해 적은 연산량으로 수행됨과 동시에 우수한 안개 제거 성능을 보여 실시간성이 요구되는 기기를 포함한 다양한 분야에 적용될 수 있음을 확인할 수 있다.

This paper presents low-complexity single-image dehazing to enhance the visibility of outdoor images that are susceptible to degradation due to weather and environmental conditions, and applies it to various devices. The conventional methods involve refinement of coarse transmission with high computational complexity and extensive memory requirements. But the proposed transmission estimation method includes excellent edge-preserving performance from comparison of the pixel-based dark channel and the patch-based dark channel in the vicinity of edges, and transmission can be estimated with low complexity since no refinement is required. Moreover, it is possible to accurately estimate transmissions and adaptively remove haze according to the characteristics of the images via prediction of the atmospheric light for each pixel using joint bright and dark channel (JBDC). Comprehensive experiments on various hazy images show that the proposed method exhibits reduced computational complexity and excellent dehazing performance, compared to the existing methods; thus, it can be applied to various fields including real-time devices.

In this paper, we describe an algorithm for extracting depth information from a single image based on CNN. When acquiring three-dimensional information from a single two-dimensional image using a deep-learning technique, it is difficult to accurately predict the edge portion of the depth image because it is a part where the depth changes abruptly. in this paper, we introduce the concept of pixel-wise confidence to take advantage of these characteristics. We propose an algorithm that estimates depth information from a highly reliable flat part and propagates it to the edge part to improve the accuracy of depth estimation.