Earticle

본 논문은 영상 속 객체와 배경의 컬러 특징을 이용한 주요 객체의 자동 추출 방법에 관한 연구이다. 인간 이 객체를 판단할 때에는 배경과 객체의 색상 차이를 이용하는데 이러한 요소를 객체 추출 방법에 적용시키기 위해 서는 배경과 객체의 색차를 강조하여야 한다. 따라서 본 논문에서는 원 RGB 영상을 인간의 시각 시스템과 유사한 HSV 색 공간으로 변환하고 각기 다른 분포도의 메디안 필터를 적용한 두 개의 영상을 생성한 뒤 두 개의 메디안 필 터가 적용된 영상들을 합산하였고 데이터 군집화 방법인 Mean Shift 알고리즘을 적용하여 색상 특징을 그룹화 하였 다. 마지막으로 이진화 작업을 위하여 영상의 채널 수를 3 채널에서 1 채널로 정규화 한 뒤 영상 내 픽셀들의 평균 값을 임계값으로 이용하는 이진화 방법으로 객체 지도 영상을 생성하였고 주요 객체를 추출하였다.

This paper is a study on an object extraction method which using color features of an object and background in the image. A human recognizes an object through the color difference of object and background in the image. So we must to emphasize the color's difference that apply to extraction result in this image. Therefore, we have converted to HSV color images which similar to human visual system from original RGB images, and have created two each other images that applied Median Filter and we merged two Median filtered images. And we have applied the Mean Shift algorithm which a data clustering method for clustering color features. Finally, we have normalized 3 image channels to 1 image channel for binarization process. And we have created object map through the binarization which using average value of whole pixels as a threshold. Then, have extracted major object from original image use that object map.

연구목적: 2022년 산업재해 현황 부가통계에서 건설업 사망사고자 현황을 보면 건설업 전체 사망사고자 의 27.8%가 건설장비로 인해 발생하고 있다. 현장 대형화, 고층화 등으로 발생하는 순회 및 점검의 한계를 극복하기 위해 컴퓨터 비전 기술을 활용해 건설장비를 추출할 수 있는 모델을 구축하고 해당 모델의 정확 도 및 현장 적용성에 대해 분석하고자 한다. 연구방법: 본 연구에서는 건설장비 중 굴착기, 덤프트럭, 이동 식 크레인의 이미지 데이터를 딥러닝 학습시킨 뒤 학습 결과를 평가 및 분석하고 건설현장에 적용하여 분 석한다. 연구결과: ‘A’ 현장에서는 굴착기 및 덤프트럭의 객체를 추출하였으며, 평균 추출 정확도는 굴착 기 81.42%, 덤프트럭 78.23%를 나타냈다. ‘B’ 현장의 이동식 크레인은 78.14%의 평균 정확도를 보여줬다. 결론: 현장 안전관리의 효율성이 증가할 수 있고, 재해발생 위험요인을 최소화 할 수 있을것이라 본다. 또 한, 본 연구를 기반으로 건설현장에 스마트 건설기술 도입에 관한 기초적인 자료로 활용이 가능하다.

Purpose: Looking at the status of fatal accidents in the construction industry in the 2022 Industrial Accident Status Supplementary Statistics, 27.8% of all fatal accidents in the construction industry are caused by construction equipment. In order to overcome the limitations of tours and inspections caused by the enlargement of sites and high-rise buildings, we plan to build a model that can extract construction equipment using computer vision technology and analyze the model's accuracy and field applicability. Method: In this study, deep learning is used to learn image data from excavators, dump trucks, and mobile cranes among construction equipment, and then the learning results are evaluated and analyzed and applied to construction sites. Result: At site ‘A’, objects of excavators and dump trucks were extracted, and the average extraction accuracy was 81.42% for excavators and 78.23% for dump trucks. The mobile crane at site ‘B’ showed an average accuracy of 78.14%. Conclusion: It is believed that the efficiency of on-site safety management can be increased and the risk factors for disaster occurrence can be minimized. In addition, based on this study, it can be used as basic data on the introduction of smart construction technology at construction sites.

본 논문에서는 기존의 그랩컷 알고리즘을 이용한 객체 검출의 정확도를 향상시키기 위하여 샤프닝 필터를 이용한 영상의 화질을 개선하는 방법을 제안한다. 그랩컷 알고리즘은 사각 윈도우 범위 내에서 객체 추출에 뛰어난 성능을 보이지만, 객체와 배경의 구분이 뚜렷하지 않은 영상에서는 성능이 떨어지는 단점이 있다. 따라서 본 논문에서는 히스토그램 평활화를 통해 밝기 및 선명도를 보강하고, 샤프닝 필터를 이용하여 객체의 경계를 강화하여 기존의 그랩컷 알고리즘보다 객체와 배경의 색상이 비슷한 영상에서 향상된 객체 추출 결과를 보인다. 개선된 그랩컷 알고리즘을 토대로 문자인식, 실시간 객체추적 등 영상처리 융합 기술에서 향상된 결과를 얻을 수 있다.

In this paper, we proposed image enhancement method using sharpening filter for improving the accuracy of object detection using the existing Grabcut algorithm. GrabCut algorithm is the excellent performance extracting an object within a rectangular window range, but it has the drawback of the inferior performance in image with no clear distinction between background and objects. So, in this paper, reinforcing the brightness and clarity through histogram equalization, and tightening the border of the object using the sharpening filter look better than that extracted result of existing GrabCut algorithm in a similar image of the object and the background. Based on improved Grabcut algorithm, it is possible to obtain an improved result in the image processing convergence technique of character recognition, real-time object tracking and so on.

In this paper, we propose a simple and efficient retrieval technique using the feature value of the corner region, which is one of the shape information attributes of images. The proposed algorithm extracts the edges and corner points of the image and rearranges the feature values of the corner regions doubly, and then measures the similarity with the image in the database using the correlation of these feature values as the feature vector. The proposed algorithm is confirmed to be more robust to rotation and size change than the conventional image retrieval method using the corner point.

In autonomous driving, environmental perception and decision-making are technologies that acquire signals based on various sensors and generate information that enables obstacle avoidance, emergency stops, and path planning. Such environmental perception technologies are limited in that they depend on expensive sensors such as LIDAR and RADAR, so research on environmental perception technologies using only cameras is actively being conducted. Studies predicting traffic accidents based on dashcam footage have also been performed as part of such research. This is challenging because only limited forward-looking footage can be acquired, and the surrounding environment is dynamic and changes quickly, making analysis difficult. Existing research has focused on learning the spatialtemporal feature representation to solve these problems. This paper extracts the trajectories of the ego-vehicle and surrounding vehicles using Visual SLAM and Multiple Object Tracking algorithms and uses them as inputs to the graph convolutional neural network(GCN) to learn the spatialtemporal feature representation. In addition, the features learned through the GCN are used as inputs to a bayesian neural network(BNN) to predict the probability of accidents, and its ability to predict accidents in advance has been verified by comparison with existing studies.

최근에 미국, 영국, 일본에서 폭발물, 독성 화학물 등에 의한 테러가 공공장소에서 시도되고 있다. 위험물을 공공장소에서 두고 가는 방식은 탐지하기 어려운 방법 중에 하나로 인식되고 있다. 공공장소에는 곳곳에 카메라가 영상 을 녹화하고 있지만, 그 영상을 사람이 일일이 모니터링 하는 것은 쉽지 않은 일이다. 최근에는 자동으로 영상을 분석하 는 지능형 소프트웨어를 유기물 탐지에 이용하고 있다. Lin 등의 방식은 비교적 높은 유기물 탐지율을 보이고 있으나, 단기 배경 영상의 특성으로 유기물에 관련한 픽셀의 수가 시간이 지날수록 급격히 감소하는 경향이 있어 그 형상 정보 를 얻기가 어렵다. 본 논문에서는 면적의 변화 특성을 분석함으로써 유기물의 형태를 성공적으로 추출하기 위한 새로운 기법을 제안한다. 제안한 방식에 대해 실험을 한 결과 선행 연구보다 형태 추출에서 우수한 성능을 보인다.

Recently the terrors have been attempted in the public places of the nations such as United states, England and Japan by explosive things, toxic materials and so on. It is understood that the method in which dangerous objects are put in public places is one of the difficult types in detection. While there are the cameras recording videos for many spots in public places, it is very hard for the security personnel to monitor every videos. Nowadays the smart softwares which can analyzing videos automatically are utilized to detect abandoned objects. The method by Lin et al. shows comparatively high detection rates for abandoned objects but it is not easy to obtain the shape information because there is a tendency that the number of the pixels decreases abruptly along the time goes due to the characteristics of short-term background images. In this research a novel method is proposed to successfully extract the shape of the abandoned object by analysing the characteristics of area variation. The experiment results show that the proposed method has better performance in extracting shape information in comparison with the precedent approach.

기존의 GrabCut 알고리즘은 자동 객체 추출이 아닌 사용자가 객체 영역에 사각형 윈도우를 설정해야하는 알고리즘이다. 본 논문에서는 자동 시스템으로 변환하기 위해 인간의 시각 시스템을 기반으로 영상에서 가장 눈에 띄는 영역을 탐지하는 방법을 연구하였다. 주의 시각 영역인 Saliency Map을 생성하기 위해서 인간이 색채를 감지하는 ‘적/녹’ ‘황/청’의 대립색설을 기반으로 하는 Lab 색공간을 이용하여 생성한다. 생성된 Saliency Map을 주파수 공간으로 변환하여 저주파 영역에 국부적인 경계를 나타내고 경계를 탐지해내어 Saliency Point를 생성한다. 이렇게 생성된 Saliency Point의 좌표 값을 이용하여 윈도우를 자동으로 생성한 후 GrabCut 알고리즘을 기반으로 객체를 추출하였다. 다양한 영상에 제안한 알고리즘을 적용한 결과 객체 영역에 자동으로 윈도우가 생성되었고 객체가 추출되었다.

Conventional GrabCut algorithm is semi-automatic algorithm that user must be set rectangle window surrounds the object. This paper studied automatic object detection to solve these problem by detecting salient region based on Human Visual System. Saliency map is computed using Lab color space which is based on color opposing theory of 'red-green' and 'blue-yellow'. Then Saliency Points are computed from the boundaries of Low-Frequency region that are extracted from Saliency Map. Finally, Rectangle windows are obtained from coordinate value of Saliency Points and these windows are used in GrabCut algorithm to extract objects. Through various experiments, the proposed algorithm computing rectangle windows of salient region and extracting objects has been proved

Kidnapping is a crime that can have disastrous results for the victim and their family. It is important to develop effective systems to detect and prevent such incidients. This paper proposes a Kidnapping Detection Systems that uses Real-Time Object Detection YOLO-v7, and Skeleton Extraction module AlphaPose to detect and track potential kidnapping event in real-time. The system utilizes a number of surveillance cameras that are already installed in Korea. It employs surveillance camera system as an edge module and a GPU system as a server module. By performing deep detection only when there is a high likelihood of a kidnapping event at the edge device, we can reduce inference costs. We have also built a dataset by recording simulated kidnapping scenarios, which can serve as a substitute for actual kidnapping events. Based on our dataset, we achieved an accuracy of 90.3% on the test set using a rule-based approach that considers the angle of the legs and occlusion with people and a car. Our system shows a promising solution for enhancing public safety and preventing crimes.

Human object extraction from infrared image has broad applications, and has become an active research area in image processing community. Combined with chaos differential evolution (CDE) algorithm and morphological operators, a novel infrared human target extraction method is proposed based on nonextensive fuzzy entropy. Firstly, the image was transformed into a fuzzy domain by fuzzy membership function, and the image nonextensive fuzzy entropy was constructed. Then, the image was segmented by thresholding based on the maximum entropy principle and the pseudoadditivity rule of nonextensive entropy. In order to reduce the search time of optimal threshold selection, the CDE algorithm was presented. Finally, the object was extracted using morphological operators to denoise, fill cavity on the threshold segmented image. Experimental results show that the proposed method is efficient and requires less computation time.

Currently product reviews information has become the focus of increasing concern to these comments often containing more information on consumer evaluation of various attributes of goods, and therefore the evaluation of mining comment object has become a hot topic. This paper presents a method called evaluation object extraction based on Bootstrapping in Chinese commodity comments, this method first comment summed up a lot of information on the properties of 10 kinds of speech rule templates, and then determined the integrity of the candidate meets these templates phrases, and these have complete of candidates as a seed, extended new property template by Bootstrapping method, and finally removed in line with all the attributes of a template as a final evaluation of the object phrase, and through experiments to calculate the accuracy, harmonic mean of recall and pr

Quan, Feng & Suh, Jinhee. 2014. A Phase Inheritance-based Account on A’-extraction of Double Object Constructions. Korean Journal of Linguistics, 39-1, 181-204. This paper attempts to explain the A’-extraction of Double Object Constructions from a cross-linguistic perspective. Three hypotheses are adopted. Firstly, every functional head may bear an EPP feature (Chomsky 1998, Landau 2007, Saito 2011), namely the generalized EPP regardless whether it is a Phase head or not, while the Phase head bears an extra EPP feature, namely the Phase EPP. Secondly, the Phase EPP feature may be handed over to the head immediately below it. Thirdly, the Mixed Chains Constraint should be respected (Chomsky 2005). In American English type-languages, since the Phase EPP is handed over to the applicative head from the little v, only the DO is allowed to undergo A’-movement to Spec CP. In Korean type-languages, since the Phase EPP feature is retained on the little v, only the IO is allowed to undergo A’-movement to Spec CP. In Polish type-languages, since the Phase EPP feature is optionally handed over to the applicative head from the little v, either the IO or the DO can undergo A’-movement to Spec CP. (Yonsei University)

In this paper, we propose a semi-automatic method for semantic video object extraction which extracts meaningful objects from an input sequence with one correctly segmented training image. Given one correctly segmented image acquired by the user's interaction in the first frame, the proposed method automatically segments and tracks the objects in the following frames. We formulate the semantic object extraction procedure as an energy minimization problem at the fragment level instead of pixel level. The proposed energy function consists of two terms: data term and smoothness term. The data term is computed by considering patch similarity, color, and motion information. Then, the smoothness term is introduced to enforce the spatial continuity. Finally, iterated conditional modes (ICM) optimization is used to minimize energy function in a globally optimal manner. The proposed semantic video object extraction method provides faithful results for various types of image sequences.

Stereo matching technique is used in many practical fields like satellite image analysis and computer vision. In this paper, we suggest a method to extract a target object image from a complicated background. For example, human face image can be extracted from random background. This method can be applied to computer vision such as security system, dressing simulation by use of extracted human face, 3D modeling, and security system. Many researches about stereo matching have been performed. Conventional approaches can be categorized into area-based and feature-based method. In this paper, we start from area-based method and apply area tracking using scanning window. Coarse depth information is used for area merging process using area searching data. Finally, we produce a target object image.

High-resolution satellite images offer abundance information of the earth surface for remote sensing applications. The information includes geometry, texture and attribute characteristic. The pixel-based image classification can't satisfy high-resolution satellite image's classification precision and produce large data redundancy. Object-oriented information extraction not only depends on spectrum character, but also use geometry and structure information. It can provide an accessible and truly revolutionary approach. Using Beijing Spot 5 high-resolution image and object-oriented classification with the eCognition software, we accomplish the cultures' precise classification. The test areas have five culture types including water, vegetation, road, building and bare lands. We use nearest neighbor classification and appraise the overall classification accuracy. The average of five species reaches 0.90. All of maximum is 1. The standard deviation is less than 0.11. The overall accuracy can reach $95.47\%.$ This method offers a new technology for high-resolution satellite images' available applications in remote sensing culture classification.

Canine cataract is developed with aging and can cause the blindness or surgical treatment if not treated timely. The first observation must be made by pet owners but they do not have proper equipment and knowledge to see the abnormalities. In this paper, we propose an intelligent image processing method to extract canine cataract suspicious object from non-professional equipment such as ordinary digital camera and cellular phone photographs so that even casual owners of pet dog can make a pre-diagnosis of such a surgery-needed disease as soon as possible. The experiment shows that the proposed method is successful in most cases except the dog has similar colored hair to the color of cataract.

When we see an object, we usually can say what it is easily even for the case where the object isn't shown in the frontal view. However, it is difficult to believe that all views of every object we have ever seen are fully memorized in our brain. Possibly, when an object is shown, we have some typical views of the object in our brain through our past experience and reconstruct the view to recognize what the presented object is. Non-negative Matrix Factorization (NMF) is one of the methods to extract the basis images from sample data set. The prominent feature of this method is that the reconstructed image is obtained by only additions of the basis images with suitable positive weights. So NMF can be seen more biologically plausible method than any other feature extraction methods such as Vector Quantization (VQ) and principal Component Analysis (PCA). In this paper, we adopt NMF to extract the aspect features from the set of images, which consists of various views of a given object. Some experiments are shown how much well NMF can extract the aspect features than any other methods such as VQ and PCA.

3차운 물체 인식 중 오목과 볼록을 갖고 있는 물체의 인식은 대단히 어려운 문제이다. 본 논문에서는 물체의 인식을 위한 피처(Feature)의 추출 방법으로 오픈-볼(Open-Ball)을 제안한다. 이 새로운 방법은 물체의 크기, 이동고 회전에 불변성을 갖는 피처(Feature)를 생성하는 것뿐만이 아니라, 비교되는 물체를 인식하는 것을 상대적인 닮음 정도 측정으로 구현한다.

Recognition of three-dimensional objects with convexities and concavities is a hard and challenging problem. This paper presents a feature extraction method out of three-dimensional objects for the purpose of classification. This new method not only provides invariance to scale, translation, and rotation $R^3$ but also distinguishes any three-dimensional model objects with concavities and convexities by measuring a relative similarity in the information space where a set of characteristics features of objects is mapped.