• Journal of the Microelectronics and Packaging Society
• /
• v.30 no.4
• /
• pp.69-78
• /
• 2023

With the rapid growth of artificial intelligence, the demand for semiconductors is enormously increasing everywhere. To ensure the manufacturing quality and quantity simultaneously, the importance of automatic defect detection during the packaging process has been re-visited by adapting various deep learning-based methodologies into automatic packaging defect inspection. Deep learning (DL) models require a large amount of data for training, but due to the nature of the semiconductor industry where security is important, sharing and labeling of relevant data is challenging, making it difficult for model training. In this study, we propose a new framework for securing sufficient data for DL models with fewer computing resources through a divide-and-conquer approach. The proposed method divides high-resolution images into pre-defined sub-regions and assigns conditional labels to each region, then trains individual sub-regions and boundaries with boundary loss inducing the globally coherent and seamless images. Afterwards, full-size image is reconstructed by combining divided sub-regions. The experimental results show that the images obtained through this research have high efficiency, consistency, quality, and generality.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.3
• /
• pp.129-136
• /
• 2020

In this paper, we propose a thermal imagery-based object detection algorithm for low-light level nighttime surveillance system. Many features selected by Haar-like feature selection algorithm and existing Adaboost algorithm are often vulnerable to noise and problems with similar or overlapping feature set for learning samples. It also removes noise from the feature set from the surveillance image of the low-light night environment, and implements it using the lightweight extended Haar feature and adaboost learning algorithm to enable fast and efficient real-time feature selection. Experiments use extended Haar feature points to recognize non-predictive objects with motion in nighttime low-light environments. The Adaboost learning algorithm with video frame 800*600 thermal image as input is implemented with CUDA 9.0 platform for simulation. As a result, the results of object detection confirmed that the success rate was about 90% or more, and the processing speed was about 30% faster than the computational results obtained through histogram equalization operations in general images.

• Journal of KIISE:Software and Applications
• /
• v.34 no.12
• /
• pp.1075-1082
• /
• 2007

In this paper, we propose the method of multiple human action recognition on video clip. For being invariant to the change of speed or size of actions, Spatial-Temporal Pyramid method is applied. Proposed method can minimize the complexity of the procedures owing to select Motion Gradient Histogram (MGH) based on statistical approach for action representation feature. For multiple action detection, Motion Energy Image (MEI) of binary frame difference accumulations is adapted and then we detect each action of which area is represented by MGH. The action MGH should be compared with pre-learning MGH having pyramid method. As a result, recognition can be done by the analyze between action MGH and pre-learning MGH. Ten video clips are used for evaluating the proposed method. We have various experiments such as mono action, multiple action, speed and site scale-changes, comparison with previous method. As a result, we can see that proposed method is simple and efficient to recognize multiple human action with stale variations.

• Journal of Intelligence and Information Systems
• /
• v.29 no.2
• /
• pp.171-187
• /
• 2023

When a person's face is recognized through a recording device such as a low-pixel surveillance camera, it is difficult to capture the face due to low image quality. In situations where it is difficult to recognize a person's face, problems such as not being able to identify a criminal suspect or a missing person may occur. Existing studies on face recognition used refined datasets, so the performance could not be measured in various environments. Therefore, to solve the problem of poor face recognition performance in low-quality images, this paper proposes a method to generate high-quality images by performing image quality improvement on low-quality facial images considering various environments, and then improve the performance of facial feature point detection. To confirm the practical applicability of the proposed architecture, an experiment was conducted by selecting a data set in which people appear relatively small in the entire image. In addition, by choosing a facial image dataset considering the mask-wearing situation, the possibility of expanding to real problems was explored. As a result of measuring the performance of the feature point detection model by improving the image quality of the face image, it was confirmed that the face detection after improvement was enhanced by an average of 3.47 times in the case of images without a mask and 9.92 times in the case of wearing a mask. It was confirmed that the RMSE for facial feature points decreased by an average of 8.49 times when wearing a mask and by an average of 2.02 times when not wearing a mask. Therefore, it was possible to verify the applicability of the proposed method by increasing the recognition rate for facial images captured in low quality through image quality improvement.

• Journal of the Korean Society of Visualization
• /
• v.19 no.3
• /
• pp.63-68
• /
• 2021

In this study, data was augmentation through the SinGAN algorithm using small image data, and defects in rubber O-rings were detected using the random forest algorithm. Unlike the commonly used data augmentation image rotation method to solve the data imbalance problem, the data imbalance problem was solved by using the SinGAN algorithm. A study was conducted to distinguish between normal products and defective products of rubber o-ring by using the random forest algorithm. A total of 20,000 image date were divided into transit and testing datasets, and an accuracy result was obtained to distinguish 97.43% defects as a result of the test.

• Proceedings of the Korean Institute of Surface Engineering Conference
• /
• 2018.06a
• /
• pp.63-63
• /
• 2018

본 연구는 플라즈마 전해 산화 (PEO) 및 RF (Radio Frequency) 스퍼터링을 이용한 2 단계 접근법에 의해 처리 된 MgO / Ni-Cr의 고내식성 이중층 코팅을 제조하기 위해 수행되었다. 이를 위해 $100mA/cm^2$ 교류 조건에서 180 s PEO를 한 후 150W 에서 900s RF 스퍼터링을 수행 하였다. 코팅의 형태는 주사전자현미경(SEM)을 사용하여 관찰되었으며 코팅의 상조성은 X-선 회절(XRD) 및 X-선 광전자 분광법(XPS)을 사용하여 분석하였다. SEM 이미지는 스퍼터링 된 Ni-Cr이 크랙의 대부분과 미세한 미세 공극을 덮어 코팅 결함이 감소함을 보여 주었다. 따라서, 코팅 된 샘플의 거칠기 값은 스퍼터링 공정 후에 감소되었다. 단면 이미지로부터, 스퍼터링된 코팅층은 낮은 두께 때문에 거의 검출되지 않았다. EDS, XRD 및 XPS를 사용한 조성 분석은 금속 상태의 형태로 Ni 및 Cr 존재를 나타내었고 XPS에서 NiCr2O4 부동태 피막이 검출되었다. MgO / Ni-Cr 이중층 코팅의 내부식성은 MgO / Ni-Cr 이중층을 가진 샘플의 금속 원소와 비교하여 우수한 부식 특성을 나타내는 전위 역학적 분극 시험 및 전기 화학적 임피던스 분광법으로 평가 하였다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.21 no.3
• /
• pp.121-126
• /
• 2020

Sound-based machine fault diagnosis is the automatic detection of abnormal sound in the acoustic emission signals of the machines. Conventional methods of using mathematical models were difficult to diagnose machine failure due to the complexity of the industry machinery system and the existence of nonlinear factors such as noises. Therefore, we want to solve the problem of machine fault diagnosis as a deep learning-based image classification problem. In the paper, we propose a CNN-based automatic machine fault diagnosis method using Spectrogram images. The proposed method uses STFT to effectively extract feature vectors from frequencies generated by machine defects, and the feature vectors detected by STFT were converted into spectrogram images and classified by CNN by machine status. The results show that the proposed method can be effectively used not only to detect defects but also to various automatic diagnosis system based on sound.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.1
• /
• pp.117-124
• /
• 2008

We suggest CBIR(Content Based Image Retrieval) method using color and shape information. Using just one feature information may cause inaccuracy compared with using more than two feature information. Therefore many image retrieval system use many feature informations like color, shape and other features. We use two feature, HSI color information especially Hue value and CSS(Curvature Scale Space) as shape information. We search candidate image form DB which include feature information of many images. When we use two features, we could approach better result.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.2
• /
• pp.19-25
• /
• 2021

In this paper proposes a method of finding missing persons based on face-recognition technology and deep learning. In this paper, a real-time face-recognition technology was developed, which performs face verification and improves the accuracy of face identification through data fortification for face recognition and convolutional neural network(CNN)-based image learning after the pre-processing of images transmitted from a mobile device. In identifying a missing person's image using the system implemented in this paper, the model that learned both original and blur-processed data performed the best. Further, a model using the pre-learned Noisy Student outperformed the one not using the same, but it has had a limitation of producing high levels of deflection and dispersion.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.82-84
• /
• 2022

최근 자율주행에서 안전한 주행을 위해 영상 기반 다중객체 검출 기술이 활발히 연구되고 있다. 이때, 저해상도 영상은 객체 검출 단계에서 정확도가 떨어지는 한계가 있다. 본 논문에서는 이러한 문제점을 해결하기 위해 초해상화와 객체 검출을 위한 방법을 함께 사용하는 기법을 제안한다. 더 나아가 초해상화 단계에서 하나의 구분자만 사용하는 기존의 방법과 다르게 이미지 생성 과정 중간에서 추가의 구분자를 사용하여 총 두 개의 구분자를 사용하여 성능을 향상하고자 하였다. 본 논문은 한국 고속도로 교통 데이터를 사용하여 실험하였으며, 그 결과 제안된 방법의 성능이 mAP@0.5 및 F1 점수 측면에서 기존 방법보다 우수하다는 것을 확인하였다.