• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 추계학술발표대회
• /
• pp.625-627
• /
• 2022

In the construction and development of modern industrial production technology, the traditional technology management mode is faced with many problems such as low qualification rates and high application costs. In the research, an improved workpiece defect detection method based on deep learning is proposed, which can control the application cost and improve the detection efficiency of irregular defects. Based on the research of the current situation of deep learning applications, this paper uses the improved Faster R-CNN network structure model as the core detection algorithm to automatically locate and classify the defect areas of the workpiece. Firstly, the robustness of the model was improved by appropriately changing the depth and the number of channels of the backbone network, and the hyperparameters of the improved model were adjusted. Then the deformable convolution is added to improve the detection ability of irregular defects. The final experimental results show that this method's average detection accuracy (mAP) is 4.5% higher than that of other methods. The model with anchor size and aspect ratio (65,129,257,519) and (0.2,0.5,1,1) has the highest defect recognition rate, and the detection accuracy reaches 93.88%.

• Geomechanics and Engineering
• /
• 제38권1호
• /
• pp.29-44
• /
• 2024

This research investigates the effectiveness of data augmentation techniques in the automated analysis of B-scan images from ground-penetrating radar (GPR) using deep learning. In spite of the growing interest in automating GPR data analysis and advancements in deep learning for image classification and object detection, many deep learning-based GPR data analysis studies have been limited by the availability of large, diverse GPR datasets. Data augmentation techniques are widely used in deep learning to improve model performance. In this study, we applied four data augmentation techniques (geometric transformation, color-space transformation, noise injection, and applying kernel filter) to the GPR datasets obtained from a testbed. A deep learning model for GPR data analysis was developed using three models (Faster R-CNN ResNet, SSD ResNet, and EfficientDet) based on transfer learning. It was found that data augmentation significantly enhances model performance across all cases, with the mAP and AR for the Faster R-CNN ResNet model increasing by approximately 4%, achieving a maximum mAP (Intersection over Union = 0.5:1.0) of 87.5% and maximum AR of 90.5%. These results highlight the importance of data augmentation in improving the robustness and accuracy of deep learning models for GPR B-scan analysis. The enhanced detection capabilities achieved through these techniques contribute to more reliable subsurface investigations in geotechnical engineering.

• 터널과지하공간
• /
• 제31권5호
• /
• pp.374-384
• /
• 2021

본 논문에서는 암석시료의 CT 촬영 이미지상의 균열을 자동으로 탐지하는 새로운 인공지능 딥러닝 기법을 제안한다. 본 제안 기법은 2단계 딥러닝 객체인식 알고르즘인 Faster R-CNN을 기반으로 회전 가능한 경계박스(bounding box) 개념을 도입하여 알고리즘을 개조하였다. 회전 경계박스의 도입은 관심 균열 영역 밖의 배경의 불균질성 및 균열의 크기와 형태에 영향을 받는 딥러닝 객체인식기법 상의 고유한 어려움을 극복하기 위한 핵심 역할을 한다. 본 회전형 경계박스의 사용은 일반적으로 사용되는 영상 수평축과 평행한 경계박스 사용의 경우와 비교하여 긴 형태의 균열 형상 특성에 매우 잘 부합된다. 즉, 좋지않은 영향을 끼치는 경계박스 내 균열 이외 배경영역의 비율을 최소화 시킬 수 있다. 이외에도, 회전 경계박스의 추가적인 이점은 인식된 균열의 방향에 따라 회전하여 추론되는 경계박스를 통해 균열의 방향과 길이에 대한 정보를 직접적으로 얻을 수 있다. 본 제안기법의 적용성을 검증하기 위하여, 이미지상에서 매우 불균질한 화강암 시료에 인공적으로 균열을 발생시킨 다수의 암석시료 영상을 딥러닝 학습에 사용하고 추론 성능 실험을 진행하였다. 그 외에도, 동일 조건에서 사암과 셰일 암석 시료에도 적용하여 검증하였다. 결론적으로, 제안된 기법을 통해 균열 객체 인식의 평균 추론정확도(mAP)값이 0.89 정도 수준의 우수한 추론 성능을 보였으며, 기존 기법에 비해 추론된 경계박스 내 균열과 배경 영역의 비율 측면에서 배경의 비율이 획기적으로 최소화되는 유리한 추론 검증 결과를 보였다.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2022년도 제66차 하계학술대회논문집 30권2호
• /
• pp.311-312
• /
• 2022

본 논문에서는 딥러닝 기반 욕창 감지를 위한 욕창 객체 탐지를 연구한다. 객체 탐지 딥러닝 기법으로 RCNN, Fast R-CNN, Faster R-CNN, YOLO 등 다양한 기법이 존재하며, 각 모델의 특징 또한 다르다. 욕창은 단계별로 피부, 조직에 손상의 정도가 다르다. 낮은 단계의 경우 일반적인 피부색과 유사하게 나타나며, 높은 단계의 경우 근육, 뼈, 지지 조직 등의 괴사로 인해 삼출물 또는 괴사조직이 나타난다. 논문에서는 One-Stage Detection 기법인 YOLO를 기반으로 욕창 이미지 내부에서 욕창 탐지를 진행한다. 현재 보유하고 있는 이미지 데이터 수가 많지 않아 데이터 증강기법을 통해 데이터를 증강하여 학습에 활용하였다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2020년도 춘계학술발표대회
• /
• pp.574-576
• /
• 2020

최근에는 딥러닝 기술의 발달로 물체 인식 및 검출에 관한 기술들 또한 발탄하고 있다. 검출에 관한 여러 기법(Faster R-CNN, R-CNN, YOLO, SSD 등) 중 SSD는 다른 기법들과는 다르게 높은 정확도와 빠른 속도가 특징이다. 동시에 여러 detection network들도 쉽게 이용이 가능하다. 본 논문에서는 detection netowork중 Mobilenet V2 network를 이용하여 SSD와 결합해 모델을 훈련하고, TensorRT engine을 이용하여 더 빠른 속도로 검출할 수 있는 방법에 대해 논의한다. 이 방법을 통해 face detector를 만들어 여러 상황에서 쓰일 수 있도록 한다.

• 한국안전학회지
• /
• 제34권6호
• /
• pp.50-57
• /
• 2019

In this study, various types of deep learning models that have been proposed recently are classified according to data input / output types and analyzed to find the deep learning model suitable for constructing a crack detection model. First the deep learning models are classified into image classification model, object segmentation model, object detection model, and instance segmentation model. ResNet-101, DeepLab V2, Faster R-CNN, and Mask R-CNN were selected as representative deep learning model of each type. For the comparison, ResNet-101 was implemented for all the types of deep learning model as a backbone network which serves as a main feature extractor. The four types of deep learning models were trained with 500 crack images taken from real concrete structures and collected from the Internet. The four types of deep learning models showed high accuracy above 94% during the training. Comparative evaluation was conducted using 40 images taken from real concrete structures. The performance of each type of deep learning model was measured using precision and recall. In the experimental result, Mask R-CNN, an instance segmentation deep learning model showed the highest precision and recall on crack detection. Qualitative analysis also shows that Mask R-CNN could detect crack shapes most similarly to the real crack shapes.

• 한국산업융합학회 논문집
• /
• 제23권6_2호
• /
• pp.979-987
• /
• 2020

This study used the Deep Learning models used in previous studies, we selected the basic model. The selected model was selected as ZFNet among ZFNet, Googlenet and ResNet, and the object was detected using a ZFNet based FRCNN. In order to reduce the detection error rate of FRCNN, location of four types of objects detected inside the image was designed by SVM classifier and location-based filtering was applied. As simulation results, it showed similar performance to the lane marking classification method with conventional 경계 detection, with an average accuracy of about 88.8%. In addition, studies using the Linear-parabolic Model showed a processing speed of 165.65ms with a minimum resolution of 600 × 800, but in this study, the resolution was treated at about 33ms with an input resolution image of 1280 × 960, so it was possible to classify lane marking at a faster rate than the previous study by CNN-based End to End method.

• 한국측량학회지
• /
• 제39권6호
• /
• pp.409-418
• /
• 2021

항공사진 촬영량이 증가함에 따라 품질검사 자동화의 필요성이 대두되고 있다. 본 연구에서는 딥러닝 기법으로 항공사진 내 구름을 분류 또는 탐지하는 실험을 수행하였고, 또한 위성영상을 학습자료에 포함시켜 분류 및 탐지를 수행하였다. 실험에 사용한 알고리즘으로는 GoogLeNet, VGG16, Faster R-CNN과 YOLOv3을 적용하여 결과를 비교하였다. 또한 구름이 포함된 오류영상 확보의 현실적 제한을 고려하여 항공영상만 존재하는 학습 데이터세트에서 위성영상을 활용한 추가학습이 분류 및 탐지정확도에 영향을 미치는지도 분석하였다. 실험결과, 항공사진의 구름 분류와 탐지에서 각각 GoogLeNet과 YOLOv3 알고리즘이 상대적으로 우월한 정확도를 나타냈고, GoogLeNet은 구름에 대한 생산자정확도 83.8% 그리고 YOLOv3는 구름에 대한 생산자정확도 84.0%를 보여주었다. 또한, 위성영상 학습자료 추가가 항공사진 자료의 부족 시 대안으로 적용가능 함을 보여주었다.

• 한국멀티미디어학회논문지
• /
• 제22권1호
• /
• pp.44-54
• /
• 2019

Recently, image-based object detection has made great progress with the introduction of Convolutional Neural Network (CNN). Many trials such as Region-based CNN, Fast R-CNN, and Faster R-CNN, have been proposed for achieving better performance in object detection. YOLO has showed the best performance under consideration of both accuracy and computational complexity. However, these data-driven detection methods including YOLO have the fundamental problem is that they can not guarantee the good performance without a large number of training database. In this paper, we propose a data sampling method using CycleGAN to solve this problem, which can convert styles while retaining the characteristics of a given input image. We will generate the insufficient data samples for training more robust object detection without efforts of collecting more database. We make extensive experimental results using the day-time and night-time road images and we validate the proposed method can improve the object detection accuracy of the night-time without training night-time object databases, because we converts the day-time training images into the synthesized night-time images and we train the detection model with the real day-time images and the synthesized night-time images.

• 한국멀티미디어학회논문지
• /
• 제22권12호
• /
• pp.1415-1429
• /
• 2019

Nowadays object tracking process becoming one of the most challenging task in Computer Vision filed. A CSR-DCF (channel spatial reliability-discriminative correlation filter) tracking algorithm have been proposed on recent tracking benchmark that could achieve stat-of-the-art performance where channel spatial reliability concepts to DCF tracking and provide a novel learning algorithm for its efficient and seamless integration in the filter update and the tracking process with only two simple standard features, HoGs and Color names. However, there are some cases where this method cannot track properly, like overlapping, occlusions, motion blur, changing appearance, environmental variations and so on. To overcome that kind of complications a new modified version of CSR-DCF algorithm has been proposed by integrating deep learning based object detection and CSRT tracker which implemented in OpenCV library. As an object detection model, according to the comparable result of object detection methods and by reason of high efficiency and celerity of Faster RCNN (Region-based Convolutional Neural Network) has been used, and combined with CSRT tracker, which demonstrated outstanding real-time detection and tracking performance. The results indicate that the trained object detection model integration with tracking algorithm gives better outcomes rather than using tracking algorithm or filter itself.