• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.3981-4004
• /
• 2019

This paper proposes a novel method for locating objects in real space from a single remote image and measuring actual distances between them by automatic detection and perspective transformation. The dimensions of the real space are known in advance. First, the corner points of the interested region are detected from an image using deep learning. Then, based on the corner points, the region of interest (ROI) is extracted and made proportional to real space by applying warp-perspective transformation. Finally, the objects are detected and mapped to the real-world location. Removing distortion from the image using camera calibration improves the accuracy in most of the cases. The deep learning framework Darknet is used for detection, and necessary modifications are made to integrate perspective transformation, camera calibration, un-distortion, etc. Experiments are performed with two types of cameras, one with barrel and the other with pincushion distortions. The results show that the difference between calculated distances and measured on real space with measurement tapes are very small; approximately 1 cm on an average. Furthermore, automatic corner detection allows the system to be used with any type of camera that has a fixed pose or in motion; using more points significantly enhances the accuracy of real-world mapping even without camera calibration. Perspective transformation also increases the object detection efficiency by making unified sizes of all objects.

• 디지털산업정보학회논문지
• /
• 제18권4호
• /
• pp.31-40
• /
• 2022

YOLOv4 can be used for detecting parking vehicles in order to check a vehicle in out-door parking space. YOLOv4 has 9 anchor boxes in each of 13x13 grid cells for detecting a bounding box of object. Because anchor boxes are allocated based on each cell, there can be existed small observational error for detecting real objects due to the distance between neighboring cells. In this paper, we proposed YOLOv4 grid cell shift algorithm for improving the out-door parking vehicle detection accuracy. In order to get more chance for trying to object detection by reducing the errors between anchor boxes and real objects, grid cells over image can be shifted to vertical, horizontal or diagonal directions after YOLOv4 basic detection process. The experimental results show that a combined algorithm of a custom trained YOLOv4 and a cell shift algorithm has 96.6% detection accuracy compare to 94.6% of a custom trained YOLOv4 only for out door parking vehicle images.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.2010-2014
• /
• 2005

Object detection and parameter estimation in point cloud data is a relevant subject to robotics, reverse engineering, computer vision, and sport mechanics. In this paper a software is presented for fully-automatic object detection and parameter estimation in unordered, incomplete and error-contaminated point cloud with a large number of data points. The software consists of three algorithmic modules each for object identification, point segmentation, and model fitting. The newly developed algorithms for orthogonal distance fitting (ODF) play a fundamental role in each of the three modules. The ODF algorithms estimate the model parameters by minimizing the square sum of the shortest distances between the model feature and the measurement points. Curvature analysis of the local quadric surfaces fitted to small patches of point cloud provides the necessary seed information for automatic model selection, point segmentation, and model fitting. The performance of the software on a variety of point cloud data will be demonstrated live.

• 한국정보전자통신기술학회논문지
• /
• 제10권4호
• /
• pp.318-324
• /
• 2017

본 항공기에 스테레오 카메라를 장착하여 영상 기반의 비행 객체 탐지 및 탐지된 객체의 3차원 위치를 추정하는 방법을 제시하였다. 구름 사이에 존재할 수 있는 원거리의 작은 객체를 탐지하기 위한 방법으로 PCT 기반의 Saliency Map을 생성하여 이용하였으며, 이렇게 탐지된 객체는 좌우 스테레오 영상에서 매칭을 수행하여 스테레오 시차(Disparity)를 추출하였다. 정확한 Disparity를 추출하기 위하여 비용집적(Cost Aggregation) 영역을 탐지 객체에 맞추어 가변되도록 가변 영역으로 사용하였으며, 본 논문에서는 Saliency Map에서 객체의 존재 영역으로 검출된 결과를 사용하였다. 좀 더 정밀한 Disparity를 추출하기 위하여 Sub-pixel interpolation 기법을 사용하여 Sub-pixel 레벨의 실수형 Disparity를 추출하였다. 또한 이에 카메라 파라미터를 적용하여 실제 탐지된 비행 객체의 3차원 공간 좌표를 생성하여 객체의 공간위치를 추정하는 방법을 제시하였다. 이는 향후 자율비행체의 영상기반 객체 탐지 및 충돌방지 시스템에 활용될 수 있을 것으로 기대된다.

• 반도체디스플레이기술학회지
• /
• 제19권2호
• /
• pp.11-15
• /
• 2020

In a typical initial setup of a PCB component inspection system, operators should manually input various information such as category, position, and inspection area for each component to be inspected, thus causing much inconvenience and longer setup time. Although there are many deep learning based object detectors, RetinaNet is regarded as one of best object detectors currently available. In this paper, a method using an extended RetinaNet is proposed that automatically detects its component category and position for each component mounted on PCBs from a high-resolution color input image. We extended the basic RetinaNet feature pyramid network by adding a feature pyramid layer having higher spatial resolution to the basic feature pyramid. It was demonstrated by experiments that the extended RetinaNet can detect successfully very small components that could be missed by the basic RetinaNet. Using the proposed method could enable automatic generation of inspection areas, thus considerably reducing the setup time of PCB component inspection systems.

• 디지털융복합연구
• /
• 제19권5호
• /
• pp.307-313
• /
• 2021

본 논문에서는 기존 객체 검출 방법 대비 매개변수를 감소시킨 경량화 네트워크를 제안하였다. 현재 사용되는 검출 모델의 경우 정확도 향상을 위해 네트워크 복잡도를 크게 늘렸다. 따라서, 제안하는 네트워크는 EfficientNet을 특징 추출 네트워크로 사용하였으며, 후속 레이어는 저수준 세부 특징과 고수준의 의미론적 특징을 활용하기 위해 피라미드 구조로 형성하였다. 피라미드 구조 사이에 attention process를 적용하여 예측에 불필요한 노이즈를 억제하였다. 네트워크의 모든 연산 과정은 depth-wise 및 point-wise 컨볼루션으로 대체하여 연산량을 최소화하였다. 제안하는 네트워크는 PASCAL VOC 데이터셋으로 학습 및 평가하였다. 실험을 통해 융합된 특징은 정제 과정을 거쳐 다양한 객체에 대해 견고한 특성을 보였다. CNN 기반 검출 모델과 비교하였을 때 적은 연산량으로 검출 정확도가 향상되었다. 향후 연구로 객체의 크기에 맞게 앵커의 비율을 조절할 필요성이 사료된다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권12호
• /
• pp.4439-4455
• /
• 2021

Compared with two-stage object detection algorithms, one-stage algorithms provide a better trade-off between real-time performance and accuracy. However, these methods treat the intermediate features equally, which lacks the flexibility to emphasize meaningful information for classification and location. Besides, they ignore the interaction of contextual information from different scales, which is important for medium and small objects detection. To tackle these problems, we propose an image pyramid network based on dual attention mechanism (DAIPNet), which builds an image pyramid to enrich the spatial information while emphasizing multi-scale informative features based on dual attention mechanisms for one-stage object detection. Our framework utilizes a pre-trained backbone as standard detection network, where the designed image pyramid network (IPN) is used as auxiliary network to provide complementary information. Here, the dual attention mechanism is composed of the adaptive feature fusion module (AFFM) and the progressive attention fusion module (PAFM). AFFM is designed to automatically pay attention to the feature maps with different importance from the backbone and auxiliary network, while PAFM is utilized to adaptively learn the channel attentive information in the context transfer process. Furthermore, in the IPN, we build an image pyramid to extract scale-wise features from downsampled images of different scales, where the features are further fused at different states to enrich scale-wise information and learn more comprehensive feature representations. Experimental results are shown on MS COCO dataset. Our proposed detector with a 300 × 300 input achieves superior performance of 32.6% mAP on the MS COCO test-dev compared with state-of-the-art methods.

• 대한원격탐사학회지
• /
• 제39권5_3호
• /
• pp.909-919
• /
• 2023

Benthic marine invertebrates, the invertebrates living on the bottom of the ocean, are an essential component of the marine ecosystem, but excessive reproduction of invertebrate grazers or pirate creatures can cause damage to the coastal fishery ecosystem. In this study, we compared and evaluated You Only Look Once Version 7 (YOLOv7), the most widely used deep learning model for real-time object detection, and detection tansformer (DETR), a transformer-based model, using underwater images for benthic marine invertebratesin the coasts of South Korea. YOLOv7 showed a mean average precision at 0.5 (mAP@0.5) of 0.899, and DETR showed an mAP@0.5 of 0.862, which implies that YOLOv7 is more appropriate for object detection of various sizes. This is because YOLOv7 generates the bounding boxes at multiple scales that can help detect small objects. Both models had a processing speed of more than 30 frames persecond (FPS),so it is expected that real-time object detection from the images provided by divers and underwater drones will be possible. The proposed method can be used to prevent and restore damage to coastal fisheries ecosystems, such as rescuing invertebrate grazers and creating sea forests to prevent ocean desertification.

• 한국방사선학회논문지
• /
• 제18권1호
• /
• pp.21-26
• /
• 2024

You only look once v5 (YOLOv5)는 객체 검출 과정에 우수한 성능을 보이고 있는 딥러닝 모델 중 하나다. 그러므로 본 연구의 목적은 양전차방출단층촬영 팬텀 영상에서 다양한 하이퍼 파라미터에 따른 YOLOv5 모델의 성능을 평가했다. 데이터 세트는 500장의 QIN PET segmentation challenge로부터 제공되는 오픈 소스를 사용하였으며, LabelImg 소프트웨어를 사용하여 경계박스를 설정했다. 학습의 적용된 하이퍼파라미터는 최적화 함수 SDG, Adam, AdamW, 활성화 함수 SiLu, LeakyRelu, Mish, Hardwish와 YOLOv5 모델 크기에 따라 nano, small, large, xlarge다. 학습성능을 평가하기 위한 정량적 분석방법으로 Intersection of union (IOU)를 사용하였다. 결과적으로, AdmaW의 최적화 함수, Hardwish의 활성화 함수, nano 크기에서 우수한 객체 검출성능을 보였다. 결론적으로 핵의학 영상에서의 객체 검출 성능에 대한 YOLOV5 모델의 유용성을 확인하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권4호
• /
• pp.1795-1811
• /
• 2019

In order to achieve rapid and accurate detection of vehicle objects in complex traffic conditions, we propose a novel vehicle detection method. Firstly, more contextual and small-object vehicle information can be obtained by our Joint Feature Network (JFN). Secondly, our Evolved Region Proposal Network (EPRN) generates initial anchor boxes by adding an improved version of the region proposal network in this network, and at the same time filters out a large number of false vehicle boxes by soft-Non Maximum Suppression (NMS). Then, our Mask Network (MaskN) generates an example that includes the vehicle occlusion, the generator and discriminator can learn from each other in order to further improve the vehicle object detection capability. Finally, these candidate vehicle detection boxes are optimized to obtain the final vehicle detection boxes by the Fine-Tuning Network(FTN). Through the evaluation experiment on the DETRAC benchmark dataset, we find that in terms of mAP, our method exceeds Faster-RCNN by 11.15%, YOLO by 11.88%, and EB by 1.64%. Besides, our algorithm also has achieved top2 comaring with MS-CNN, YOLO-v3, RefineNet, RetinaNet, Faster-rcnn, DSSD and YOLO-v2 of vehicle category in KITTI dataset.