• 로봇학회논문지
• /
• 제5권2호
• /
• pp.85-92
• /
• 2010

In this paper, we propose a digital image stabilization technique using edge detection and Lucas-Kanade optical flow in order to minimize the motion of the shaken image. The accuracy of motion estimation based on block matching technique depends on the size of search window, which results in long calculation time. Therefore it is not applicable to real-time system. In addition, since the size of vector depends on that of block, it is difficult to estimate the motion which is bigger than the block size. The proposed method extracts the trust region using edge detection, to estimate the motion of some critical points in trust region based on Lucas-Kanade optical flow algorithm. The experimental results show that the proposed method stabilizes the shaking of motion image effectively in real time.

• Journal of Multimedia Information System
• /
• 제9권3호
• /
• pp.183-190
• /
• 2022

Image-based 3D object detection is one of the important and difficult problems in autonomous driving and robotics, and aims to find and represent the location, dimension and orientation of the object of interest. It generates three dimensional (3D) bounding boxes with only 2D images obtained from cameras, so there is no need for devices that provide accurate depth information such as LiDAR or Radar. Image-based methods can be divided into three main categories: monocular, stereo, and multi-view 3D object detection. In this paper, we investigate the recent state-of-the-art models of the above three categories. In the multi-view 3D object detection, which appeared together with the release of the new benchmark datasets, NuScenes and Waymo, we discuss the differences from the existing monocular and stereo methods. Also, we analyze their performance and discuss the advantages and disadvantages of them. Finally, we conclude the remaining challenges and a future direction in this field.

• 전자공학회논문지CI
• /
• 제39권1호
• /
• pp.67-74
• /
• 2002

멀티미디어 기기의 발전과 더불어 다양한 매체로부터 다양한 종류의 영상, 오디오, 텍스트 등의 정보가 난무하고, 이들 정보를 사용자의 요구에 따라 효과적으로 검색·관리를 위한 연구가 활발히 진행되어지고 있다. 실시간 연속 영상에서 내용이 연결되는 부분과 장면전환 등에 의해 내용이 바뀌는 부분을 자동 검출 가능하다면 적은 량의 데이터 내용 표현으로 영상 검색의 효율성을 증대시키는 효과를 가져 올 수 있을 것이다. 본 논문에서는 영상의 특성에 따른 실시간 Cut detection 기술을 제안하고 이 방법의 성능을 다양한 영상 데이터를 바탕으로 정확성 평가를 하였다. 그 결과 영상데이터의 컬러 특성에 관한 통계적인 특성 정보를 필요로 하는 기존의 컬러 히스토그램 방식과는 달리 본 방식은 각 프레임 영상의 색상 분포의 변화분에 의존하므로 어떤 종류의 영상 패턴에도 적용 가능한 robust한 방식이며, 실시간 입력영상의 cut detection 이 가능한 이점이 있음을 확인할 수 있었다.

• 한국도로학회논문집
• /
• 제15권5호
• /
• pp.177-185
• /
• 2013

PURPOSES : This study attempts to design and establish the road surface condition detection system by using the image processing that is expected to help implement the low-cost and high-efficiency road information detection system by examining technology trends in the field of road surface condition information detection and related case studies. METHODS : Adapted visual information collecting method(setting a stereo camera outside of the vehicle) and visual information algorithm(transform a Wavelet Transform, using the K-means clustering) Experiments and Analysis on Real-road, just as four states(Dry, Wet, Snow, Ice). RESULTS : Test results showed that detection rate of 95% or more was found under the wet road surface, and the detection rate of 85% or more in snowy road surface. However, the low detection rate of 30% was found under the icy road surface. CONCLUSIONS : As a method to improve the detection rate of the mobile road surface condition information detection system developed in this study, more accurate phase analysis in the image processing process was needed. If periodic synchronization through automatic settings of the camera according to weather or ambient light was not made at the time of image acquisition, a significant change in the values of polarization coefficients occurs.

• 대한임베디드공학회논문지
• /
• 제17권1호
• /
• pp.41-49
• /
• 2022

This paper proposes a novel lane detection algorithm based on inverse perspective transformation and machine learning in lightweight embedded system. The inverse perspective transformation method is presented for obtaining a bird's-eye view of the scene from a perspective image to remove perspective effects. This method requires only the internal and external parameters of the camera without a homography matrix with 8 degrees of freedom (DoF) that maps the points in one image to the corresponding points in the other image. To improve the accuracy and speed of lane detection in complex road environments, machine learning algorithm that has passed the first classifier is used. Before using machine learning, we apply a meaningful first classifier to the lane detection to improve the detection speed. The first classifier is applied in the bird's-eye view image to determine lane regions. A lane region passed the first classifier is detected more accurately through machine learning. The system has been tested through the driving video of the vehicle in embedded system. The experimental results show that the proposed method works well in various road environments and meet the real-time requirements. As a result, its lane detection speed is about 3.85 times faster than edge-based lane detection, and its detection accuracy is better than edge-based lane detection.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2006년도 Proceedings of ISRS 2006 PORSEC Volume I
• /
• pp.260-262
• /
• 2006

One of future remote sensing techniques for transportation application is vehicle detection from the space, which could be the basis of measuring traffic volume and recognizing traffic condition in the future. This paper introduces an approach to vehicle detection using image object segmentation approach. The object-oriented image processing is particularly beneficial to high-resolution image classification of urban area, which suffers from noisy components in general. The project site was Dae-Jeon metropolitan area and a set of true color aerial images at 10cm resolution was used for the test. Authors investigated a variety of parameters such as scale, color, and shape and produced a customized solution for vehicle detection, which is based on a knowledge-based hierarchical model in the environment of eCognition. The highest tumbling block of the vehicle detection in the given data sets was to discriminate vehicles in dark color from new black asphalt pavement. Except for the cases, the overall accuracy was over 90%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권9호
• /
• pp.3751-3770
• /
• 2015

Many sorts of image processing software facilitate image editing and also generate a great number of doctored images. Forensic technology emerges to detect the unintentional or malicious image operations. Most of forensic methods focus on the detection of single operations. However, a series of operations may be used to sequentially manipulate an image, which makes the operation detection problem complex. Forensic investigators always want to know as much exhaustive information about a suspicious image's entire processing history as possible. The detection of the operation chain, consisting of a series of operations, is a significant and challenging problem in the research field of forensics. In this paper, based on the histogram distribution uniformity of a manipulated image, we propose an operation chain detection scheme to identify histogram equalization (HE) followed by the dither-like operation (DLO). Two histogram features and a local spatial feature are utilized to further determine which DLO may have been applied. Both theoretical analysis and experimental results verify the effectiveness of our proposed scheme for both global and local scenarios.

• 한국멀티미디어학회논문지
• /
• 제21권8호
• /
• pp.888-896
• /
• 2018

In this paper, we report detection and recognition of vehicle logo from images captured from street CCTV. Image data includes both the front and rear view of the vehicles. The proposed method is a two-step process which combines image preprocessing and faster region-based convolutional neural network (R-CNN) for logo recognition. Without preprocessing, faster R-CNN accuracy is high only if the image quality is good. The proposed system is focusing on street CCTV camera where image quality is different from a front facing camera. Using perspective transformation the top view images are transformed into front view images. In this system, the detection and accuracy are much higher as compared to the existing algorithm. As a result of the experiment, on day data the detection and recognition rate is improved by 2% and night data, detection rate improved by 14%.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.1099-1101
• /
• 2003

Change detection technology, which discovers the change information on the surface of the earth by comparing and analyzing multi-temporal satellite images, can be usefully applied to the various fields, such as environmental inspection, urban planning, forest policy, updating of geographical information and the military usage. In this paper, we introduce a change detection system that can extract and analyze change elements from high-resolution satellite imagery as well as low- or middle-resolution satellite imagery. The developed system provides not only 7 pixelbased methods that can be used to detect change from low- or middle-resolution satellite images but also a float window concept that can be used in manual change detection from highresolution satellite images. This system enables fast access to the very large image, because it is constituted by OGC grid coverage components. Also new change detection algorithms can be easily added into this system if once they are made into grid coverage components.

• International journal of advanced smart convergence
• /
• 제10권3호
• /
• pp.163-171
• /
• 2021

Recently, Machine Learning-based visualization approaches have been proposed to combat the problem of malware detection. Unfortunately, these techniques are exposed to Adversarial examples. Adversarial examples are noises which can deceive the deep learning based malware detection network such that the malware becomes unrecognizable. To address the shortcomings of these approaches, we present Block-matching and 3D filtering (BM3D) algorithm and deep image prior based denoising technique to defend against adversarial examples on visualization-based malware detection systems. The BM3D based denoising method eliminates most of the adversarial noise. After that the deep image prior based denoising removes the remaining subtle noise. Experimental results on the MS BIG malware dataset and benign samples show that the proposed denoising based defense recovers the performance of the adversarial attacked CNN model for malware detection to some extent.