• 대한원격탐사학회지
• /
• 제34권6_3호
• /
• pp.1427-1443
• /
• 2018

객체 기반 영상 분석은 영상의 복잡도를 낮추는 동시에 영상의 특성을 유지한다는 점에서 픽셀 기반 영상 분석보다 높은 효율성과 정보 활용 가능성을 지닌다. Superpixel은 일반적인 영상 분할보다 작은 영상 단위로 영상을 과분할함으로써 영상 내의 경계를 보다 잘 유지할 수 있다. 이 가운데 SLIC(Simple linear iterative clustering) superpixel 기법은 기존의 기법들보다 높은 품질의 영상 분할 결과를 제시하는 것으로 알려져 있다. 이러한 SLIC 기법의 입력 파라미터인 superpixel의 개수는 영상 분할 결과에 큰 영향을 미침에도 이에 대한 연구는 선행 연구에서 충분히 다루어지지 않았다. 이에 본 연구에서는 KOMPSAT 영상을 이용하여 변화 탐지 활용 연구를 위한 SLIC 계열 superpixel 기법의 최적 파라미터 분석 및 변화 탐지 성능 비교를 수행하였다. 사용된 superpixel 기법은 SLIC, SLIC0(SLIC의 무변수 버전), SNIC(Simple non-iterative clustering) 의 세 가지 기법으로, $5{\times}5$(픽셀)에서 $50{\times}50$(픽셀)의 superpixel 크기 범위에 대해서 superpixel 개수를 지정하여 superpixel 분할 영상을 생성하고 변화 탐지 참조 영상에 대한 재현율을 분석하였다. 이를 통해 얻어진 최적 superpixel 크기를 바탕으로 변화를 탐지하고자 하는 두 영상의 차 영상을 분할한 후 일정 크기의 객체로 clustering하였다. 두 시기(bi-temporal) 영상으로부터 얻어진 공통된 영상경계는 전후 영상에 각각 적용함으로써 각 superpixel의 feature(Lab 색상 차이) 변화를 탐지하였다. 최종적인 변화 탐지 결과는 참조 영상을 통해 그 성능이 분석하였으며, 영상의 과분할 정도가 높지 않더라도 규칙적인 크기와 형태의 superpixel을 통해 높은 변화 탐지 성능을 달성할 수 있음을 확인하였다.

• 디지털산업정보학회논문지
• /
• 제10권4호
• /
• pp.151-160
• /
• 2014

In this paper, a comparison of superpixel characteristics based on SLIC(simple linear iterative clustering) for several color feature spaces is presented. Computer vision applications have come to rely increasingly on superpixels in recent years. Superpixel algorithms group pixels into perceptually meaningful atomic regions, which can be used to replace the rigid structure of the pixel grid. A superpixel is consist of pixels with similar features such as luminance, color, textures etc. Thus superpixels are more efficient than pixels in case of large scale image processing. Generally superpixel characteristics are described by uniformity, boundary precision and recall, compactness. However previous methods only generate superpixels a special color space but lack researches on superpixel characteristics. Therefore we present superpixel characteristics based on SLIC as known popular. In this paper, Lab, Luv, LCH, HSV, YIQ and RGB color feature spaces are used. Uniformity, compactness, boundary precision and recall are measured for comparing characteristics of superpixel. For computer simulation, Berkeley image database(BSD300) is used and Lab color space is superior to the others by the experimental results.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2011년도 추계학술대회
• /
• pp.915-917
• /
• 2011

본 논문에서는 칼라특징공간별로 슈퍼픽셀의 특성을 비교하였다. 슈퍼픽셀은 특성이 비슷한 인접 화소들을 묶어서 하나의 큰 화소로 취급하는 것으로 고속영상처리 및 인식을 위해 사용한다. 본 연구에서는 칼라특징공간별로 슈퍼픽셀을 구하여 각각의 특징을 비교하고자 한다. 비교할 특징은 슈퍼 픽셀의 중요한 특징인 밀집성(compactness)사용한다. 실험에 사용한 영상은 버클리대학교의 영상분할 데이터베이스인 BSD-300영상을 사용하여 실험하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권5호
• /
• pp.2576-2589
• /
• 2017

Co-saliency detection is a task of detecting same or similar objects in multi-scene, and has been an important preprocessing step for multi-scene image processing. However existing methods lack efficiency to match similar areas from different images. In addition, they are confined to single image detection without a unified framework to calculate co-saliency. In this paper, we propose a novel model called Superpixel Matching-Cellular Automata (SMCA). We use Hausdorff distance adjacent superpixel sets instead of single superpixel since the feature matching accuracy of single superpixel is poor. We further introduce Cellular Automata to exploit the intrinsic relevance of similar regions through interactions with neighbors in multi-scene. Extensive evaluations show that the SMCA model achieves leading performance compared to state-of-the-art methods on both efficiency and accuracy.

• 디지털산업정보학회논문지
• /
• 제11권4호
• /
• pp.89-97
• /
• 2015

In this paper, a method of color image segmentation based on DBSCAN(Density Based Spatial Clustering of Applications with Noise) using compactness of superpixels and texture information is presented. The DBSCAN algorithm can generate clusters in large data sets by looking at the local density of data samples, using only two input parameters which called minimum number of data and distance of neighborhood data. Superpixel algorithms group pixels into perceptually meaningful atomic regions, which can be used to replace the rigid structure of the pixel grid. Each superpixel is consist of pixels with similar features such as luminance, color, textures etc. Superpixels are more efficient than pixels in case of large scale image processing. In this paper, superpixels are generated by SLIC(simple linear iterative clustering) as known popular. Superpixel characteristics are described by compactness, uniformity, boundary precision and recall. The compactness is important features to depict superpixel characteristics. Each superpixel is represented by Lab color spaces, compactness and texture information. DBSCAN clustering method applied to these feature spaces to segment a color image. To evaluate the performance of the proposed method, computer simulation is carried out to several outdoor images. The experimental results show that the proposed algorithm can provide good segmentation results on various images.

• 한국멀티미디어학회논문지
• /
• 제19권3호
• /
• pp.531-538
• /
• 2016

In this paper, we propose a novel post-processing method of semantic segmentation from indoor scenes with RGBD inputs. For accurate segmentation, various post-processing methods such as superpixel from color edges or Conditional Random Field (CRF) method considering neighborhood connectivity have been used, but these methods are not efficient due to high complexity and computational cost. To solve this problem, we maximize the efficiency of post processing by using depth superpixel extracted from disparity image to handle object silhouette. Our experimental results show reasonable performances compared to previous methods in the post processing of semantic segmentation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권9호
• /
• pp.4375-4388
• /
• 2018

To solve the problem of transmission errors in stereoscopic images, this paper proposes a novel error concealment (EC) method using superpixel segmentation and adaptive disparity selection (SSADS). Our algorithm consists of two steps. The first step is disparity estimation for each pixel in a reference image. In this step, the numbers of superpixel segmentation labels of stereoscopic images are used as a new constraint for disparity matching to reduce the effect of mismatching. The second step is disparity selection for a lost block. In this step, a strategy based on boundary smoothness is proposed to adaptively select the optimal disparity which is used for error concealment. Experimental results demonstrate that compared with other methods, the proposed method has significant advantages in both objective and subjective quality assessment.

• 한국멀티미디어학회논문지
• /
• 제19권6호
• /
• pp.1003-1013
• /
• 2016

This paper proposes a framework of superpixel-based vehicle detection method using plane normal vector in disparity space. We utilize two common factors for detecting vehicles: Hypothesis Generation (HG) and Hypothesis Verification (HV). At the stage of HG, we set the regions of interest (ROI) by estimating the lane, and track them to reduce computational cost of the overall processes. The image is then divided into compact superpixels, each of which is viewed as a plane composed of the normal vector in disparity space. After that, the representative normal vector is computed at a superpixel-level, which alleviates the well-known problems of conventional color-based and depth-based approaches. Based on the assumption that the central-bottom of the input image is always on the navigable region, the road and obstacle candidates are simultaneously extracted by the plane normal vectors obtained from K-means algorithm. At the stage of HV, the separated obstacle candidates are verified by employing HOG and SVM as for a feature and classifying function, respectively. To achieve this, we trained SVM classifier by HOG features of KITTI training dataset. The experimental results demonstrate that the proposed vehicle detection system outperforms the conventional HOG-based methods qualitatively and quantitatively.

• 스마트미디어저널
• /
• 제8권2호
• /
• pp.39-45
• /
• 2019

딥러닝이 보편화되면서 예측 결과를 설명하는 연구가 중요해졌다. 최근 슈퍼픽셀에 기반한 다중스케일 결합 기법이 제안되었는데, 물체의 모양을 유지함으로써 시각적 공감이라는 장점을 제공한다. 이 기법은 예측 차이라는 원리에 기반을 두고 있으며, 슈퍼픽셀을 가리고 얻은 예측 결과와 원래 예측 결과의 차이를 보고 돌출맵을 구성한다. 본 논문은 슈퍼픽셀을 가리는 제외 연산뿐 아니라 슈퍼픽셀만 보여주는 포함 연산까지 사용하는 새로운 기법을 제안한다. 실험 결과 제안한 방법은 IoU에서 3.3%의 성능 향상을 보인다.

• 한국표면공학회지
• /
• 제55권3호
• /
• pp.164-172
• /
• 2022

Hot-dip galvanized steel(GI) is widely used throughout the industry as a corrosion resistance material. Corrosion of steel is a common phenomenon that results in the gradual degradation under various environmental conditions. Corrosion monitoring is to track the degradation progress for a long time. Corrosion on steel plate appears as discoloration and any irregularities on the surface. This study developed a quantitative evaluation method of the rust formed on GI steel plate using a superpixel-based DBSCAN clustering method and k-means clustering from the corroded area in a given image. The superpixel-based DBSCAN clustering method decrease computational costs, reaching automatic segmentation. The image color of the rusty surface was analyzed quantitatively based on HSV(Hue, Saturation, Value) color space. In addition, two segmentation methods are compared for the particular spatial region using their histograms.