• Smart Media Journal
• /
• v.9 no.2
• /
• pp.9-15
• /
• 2020

In this paper, we propose a network model composed of Multi path Encoder-Decoder branches that can recognize each instance from the image. The network has two branches, Dot branch and Segmentation branch for finding the center point of each instance and for recognizing area of the instance, respectively. In the experiment, the CVPPP dataset was studied to distinguish leaves from each other, and the center point detection branch(Dot branch) found the center points of each leaf, and the object segmentation branch(Segmentation branch) finally predicted the pixel area of each leaf corresponding to each center point. In the existing segmentation methods, there were problems of finding various sizes and positions of anchor boxes (N > 1k) for checking objects. Also, there were difficulties of estimating the number of undefined instances per image. In the proposed network, an effective method finding instances based on their center points is proposed.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.04a
• /
• pp.656-659
• /
• 2016

최근 컴퓨터 기술의 발전과 함께 임베디드 기기 또한 다양한 기능을 갖추기 시작했다. 본 연구에서는 최근 활발하게 진행되고 있는 영상센서를 사용한 임베디드 기기 등 자원이 적은 기기에서 효율적인 얼굴 추적 방식을 제안한다. 정확한 얼굴을 얻기 위하여 MB-LBP 특징을 사용한 얼굴 검출 방식을 사용했으며, 다음 영상에서 얼굴 객체 추적을 위하여 얼굴 검출시 얼굴 주변 영역(Region of Interest)을 지정하였다. 그리고 얼굴을 검출을 못하는 영상에서는 기존의 객체 추적 방식인 CAM-Shift를 사용해 객체를 추적해 객체 정보의 손실 없이 정보를 유지 할 수 있도록 하였다. 본 연구는 기존 연구와의 비교를 통하여 객체 추적 시스템의 정확성과 빠른 성능을 확인하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.11a
• /
• pp.160-162
• /
• 2018

본 논문에서는 압축 영역에서 열 영상을 이용한 딥러닝 기반의 객체 검출 방법을 제안한다. 비디오 압축 표준인 High Efficiency Video Coding(HEVC)를 이용하여 부보화된 비트스트림으로부터 Intra Prediction Mode(IPM), Prediction Unit Size(PUS), Transform Unit Size(TUS)를 추출하고 3 채널 영상으로 변환하고 객체 검출 네트워크인 YOLO 에 입력으로 넣어주어 최종적으로 객체의 위치 및 객체의 종류를 예측한다. 실험결과로써 복원된 열 영상과 검출된 결과를 주관적으로 보여줌으로써 압축영역에서 열영상을 이용한 객체 검출이 가능함을 보인다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.5
• /
• pp.274-286
• /
• 2022

Multi-object tracking has been studied for a long time under computer vision and plays a critical role in applications such as autonomous driving and driving assistance. Multi-object tracking techniques generally consist of a detector that detects objects and a tracker that tracks the detected objects. Various publicly available datasets allow us to train a detector model without much effort. However, there are relatively few publicly available datasets for training a tracker model, and configuring own tracker datasets takes a long time compared to configuring detector datasets. Hence, the detector is often developed separately with a tracker module. However, the separated tracker should be adjusted whenever the former detector model is changed. This study proposes a system that can train a model that performs detection and tracking simultaneously using only the detector training datasets. In particular, a Siam network with augmentation is used to compose the detector and tracker. Experiments are conducted on public datasets to verify that the proposed algorithm can formulate a real-time multi-object tracker comparable to the state-of-the-art tracker models.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.446-450
• /
• 2020

딥러닝 기반의 영상 분석 방법들은 많은 양의 학습 데이터가 필요하며, 학습 데이터 구축에는 많은 시간과 노력이 소요된다. 특히 객체 검출 분야의 경우 영상 내 객체의 위치, 크기, 범주 등의 정보가 모두 필요하여 학습 데이터 구축에 더 많은 어려움이 있으며, 이를 해결하기 위해 최근 이미지 합성기반 데이터 증강에 관한 연구가 활발히 진행되고 있다. 이미지 합성기반 데이터 증강 방법은 배경 영상에 객체를 합성할 때 객체와 배경 영상이 접한 영역에서 아티팩트(Artifact)가 발생하며, 이는 객체 검출 모델이 아티팩트를 객체의 특징으로 모델링하여 검출 성능이 저하되는 원인이 된다. 이러한 문제를 해결하기 위하여 본 논문에서는 양방향 필터 기반의 이미지 합성 방법을 제안하고, 단일 단계 검출의 대표적인 방법인 RetinaNet을 이용하여 이미지 합성기반 데이터 증강 방법의 성능을 분석하였다. 공개 데이터셋에 대한 실험 결과 본 논문에서 사용한 단일 검출 방법 및 데이터 증강 기법을 사용하면 더 적은 양의 증강 데이터로 기존 방법과 동일한 성능을 보여주는 것을 확인하였다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.10a
• /
• pp.894-897
• /
• 2012

We propose a effective method using the HOG (Histogram of Oriented Gradients) feature vector to track individual objects in an environment which multiple objects are moving. The proposed algorithm consists of pre-processing, object detection and object tracking. We experimented with six videos which have various trajectories and the movement. When occlusion between objects was occurred, we identified individual object by using center and predicted coordinates of moving objects. The algorithm shows 85.45% of tracking rate in the videos we experimented. We expect the proposed system is utilized in security systems which require the alalysis of the position and motion pattern of objects.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.10
• /
• pp.323-328
• /
• 2020

Recently, new approaches that significantly improve performance in object detection and recognition using deep learning technology have been proposed quickly. Of the various techniques for object detection, especially facial object detection (Faster R-CNN, R-CNN, YOLO, SSD, etc), SSD is superior in accuracy and speed to other techniques. At the same time, multiple object detection networks are also readily available. In this paper, among object detection networks, Mobilenet v2 network is used, models combined with SSDs are trained, and methods for detecting objects at a rate of four times or more than conventional performance are proposed using TensorRT engine, and the performance is verified through experiments. Facial object detector was created as an application to verify the performance of the proposed method, and its behavior and performance were tested in various situations.

• Journal of Broadcast Engineering
• /
• v.28 no.1
• /
• pp.124-131
• /
• 2023

Object detection techniques are increasingly used to obtain information on physical characteristics or situations of a specific area from remote images. The accuracy of object detection is decreased in remote sensing images with low resolution because the low resolution reduces the amount of detail that can be captured in an image. A single neural network is proposed to joint the super-resolution method and object detection method. The proposed method constructs a deep residual-based network to restore object features in low-resolution images. Moreover, the proposed method is used to improve the performance of object detection by jointing a single network with YOLOv5. The proposed method is experimentally tested using VEDAI data for low-resolution images. The results show that vehicle detection performance improved by 81.38% on mAP@0.5 for VISIBLE data.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.91-94
• /
• 2022

자율주행 자동차 개발 연구가 활발히 진행됨에 따라 객체 검출기의 성능이 중요하게 되었다. 딥러닝 기술의 발전하면서 객체 검출기의 성능도 큰 발전을 이루었다. 그에 따라 도로 위 차량 검출기의 성능도 발전하고 있으나 평상시 낮 도로상황에서 잘 동작하던 모델은 안개가 끼거나 밤 상황이 되면 제대로 동작하지 못하는 문제를 가지고 있다. 이유는 딥러닝 모델이 학습할 때 사용한 데이터셋의 정보에 따라 특정 도메인에 편향된 특성을 학습하기 때문이다. 따라서, 본 논문에서는 객체 검출 신경망에 도메인 판별기를 적용하여 이와 같은 도메인 이동 문제를 극복하는 모델을 제안한다. 모델의 성능을 Cityscapes 데이터셋과 Foggy Cityscapes 데이터셋을 사용하여 평가한 결과, 기존의 특정 도메인에서 학습한 모델보다 제안하는 모델의 검출 성능이 개선된다는 것을 확인하였다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.4
• /
• pp.43-49
• /
• 2019

The wired and wireless Internet is a useful window to easily acquire various types of media data. On the other hand, the public can easily get the media data including the object to which the personal information is exposed, which is a social problem. In this paper, we propose a method to robustly detect a target object that has exposed personal information using a learning algorithm and effectively block the detected target object area. In the proposed method, only the target object containing the personal information is detected using a neural network-based learning algorithm. Then, a grid-like mosaic is created and overlapped on the target object area detected in the previous step, thereby effectively blocking the object area containing the personal information. Experimental results show that the proposed algorithm robustly detects the object area in which personal information is exposed and effectively blocks the detected area through mosaic processing. The object blocking method presented in this paper is expected to be useful in many applications related to computer vision.