• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.7
• /
• pp.80-86
• /
• 2019

As computer performance improves, research using deep learning are being actively carried out in various fields. Recently, deep learning technology has been applying to the safety evaluation for structures. In particular, the internal blades of a turbine structure requires experienced experts and considerable time to detect surface damages because of the difficulty of separation of the blades from the structure and the dark environmental condition. This study proposes a Faster R-CNN deep learning model that can detect surface damages on the internal blades, which is one of the primary elements of the turbine structure. The deep learning model was trained using image data with dent and punch damages. The image data was also expanded using image filtering and image data generator techniques. As a result, the deep learning model showed 96.1% accuracy, 95.3% recall, and 96% precision. The value of the recall means that the proposed deep learning model could not detect the blade damages for 4.7%. The performance of the proposed damage detection system can be further improved by collecting and extending damage images in various environments, and finally it can be applicable for turbine engine maintenance.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.1
• /
• pp.191-196
• /
• 2021

In this paper, we implemented an image sensor module possible of detecting color temperature in an indoor environment. The color temperature information in the video information acquired by the image sensor was matched with a color difference illuminometer to produce an LUT. An algorithm was developed so that color temperature information according to the received RGB values can be automatically calculated. As a result of measuring the color temperature with an image sensor indoors, an accurate result of less than 5.91% was obtained compared to the reference value. It was confirmed that the uniformity of 23.5% or more was excellent compared to the color temperature measurement result using a color sensor.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.85-87
• /
• 2022

루푸스 신염을 정확히 진단하기 위해서는 신장의 침 생검을 통한 조직검사를 통해 사구체들을 찾아내고, 각각의 염증 정도를 분류해야 한다. 하지만 이에는 의료진의 많은 시간과 노력이 소요된다. 따라서 본 연구에서는 이러한 한계를 극복하기 위해 합성곱 신경망 (Convolutional neural network, CNN)에 기반한 검출 및 분할에 딥 러닝 접근법을 적용하는 YOLOv5 알고리즘을 통해 검체 이미지 내에서 사구체를 자동으로 검출해 내도록 하였다. 그리고 루푸스 신염 환자의 슬라이드 이미지에 대한 태깅 작업을 거쳐 학습을 위한 데이터와 테스트를 위한 데이터를 생성하여 학습 및 테스트에 활용하였다. 그 결과 고화질의 검체 이미지 내에서 대부분의 사구체를 0.9 이상의 높은 precision과 recall로 검출해 낼 수 있었다. 이를 통해 신장 내부의 사구체 검출을 자동화하고 추후 연구를 통해 사구체 염증 정도를 단계화 할 수 있는 발판을 마련하였다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.9
• /
• pp.1681-1688
• /
• 2017

In this paper, we implemented a high-speed signal processing hardware system for Color Line Scan Camera using FPGA and Nor-Flash. The existing hardware system mainly processed by high-speed DSP based on software and it was a method of detecting defects mainly by RGB individual logic, however we suggested defect detection hardware using RGB-HSL hardware converter, FIFO, HSL Full-Color Defect Decoder and Image Frame Buffer. The defect detection hardware is composed of hardware look-up table in converting RGB to HSL and 4K HSL Full-Color Defect Decoder with high resolution. In addition, we included an image frame for comprehensive image processing based on two dimensional image by line data accumulation instead of local image processing based on line data. As a result, we can apply the implemented system to the grain sorting machine for the sorting of peanuts effectively.

• The Journal of the Korea Contents Association
• /
• v.13 no.10
• /
• pp.54-61
• /
• 2013

This paper presents a simplified hardware based edge detection circuit which is based on an analog correlator combining with the neighbor pixels in CMOS image sensor. A pixel element of the edge detector consists of an active pixel sensor and an analog correlator circuit which connects two neighbor pixels. The edge detector shares a comparator on each column that the comparator decides an edge of the target pixel with an adjustable reference voltage. The circuit detects image edge from CIS directly that reduces area and power consumption 4 times and 20%, respectively, compared with the previous works. And also it has advantage to regulate sensitivity of the edge detection because the threshold value is able to control externally. The fabricated chip has 34% of fill factor and 0.9 ${\mu}W$ of power per a pixel under 0.18 ${\mu}m$ CMOS technology.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2018.07a
• /
• pp.325-326
• /
• 2018

본 논문에서는 PCB 공정상의 육안검사를 통한 불량 분류 방식에서 CNN을 이용한 PCB 불량 분류 방식을 제안한다. 이 방식은 육안검사의 문제점인 작업자의 숙련도에 따른 검사 효율을 자동화 검사 시스템에 의해 해결하며, 불량 위치와 종류를 결과 이미지에 표시한다. 또한 이미지 분류 결과를 모니터링할 수 있도록 시리얼 통신을 통하여 Darknet 프레임워크와 LCD를 연동하였다. 적은 량의 데이터 셋으로도 좋은 결과를 냈으며, 다양한 데이터 셋을 이용해 훈련할 시 전반적인 PCB 불량의 분류가 가능할 것으로 예상된다.

• Journal of Digital Convergence
• /
• v.10 no.6
• /
• pp.297-302
• /
• 2012

In this paper, we present a object detection and tracking method based on image context analysis. It is robust from the image variations such as complicated background, dynamic movement of the object. Image context analysis is carried out using the hybrid network of k-means and RBF. The proposed object detection employs context-driven adaptive Bayesian framework to relive the effect due to uneven object images. The proposed method used feature vector generator using 2D Haar wavelet transform and the Bayesian discriminant method in order to enhance the speed of learning. The system took less time to learn, and learning in a wide variety of data showed consistent results. After we developed the proposed method was applied to real-world environment. As a result, in the case of the object to detect pass outside expected area or other changes in the uncertain reaction showed that stable. The experimental results show that the proposed approach can achieve superior performance using various data sets to previously methods.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.41-51
• /
• 2024

In this paper, we explore the enhancement of target detection accuracy in the guided weapon using deep learning object detection on infrared (IR) images. Due to the characteristics of IR images being influenced by factors such as time and temperature, it's crucial to ensure a consistent representation of object features in various environments when training the model. A simple way to address this is by emphasizing the features of target objects and reducing noise within the infrared images through appropriate pre-processing techniques. However, in previous studies, there has not been sufficient discussion on pre-processing methods in learning deep learning models based on infrared images. In this paper, we aim to investigate the impact of image pre-processing techniques on infrared image-based training for object detection. To achieve this, we analyze the pre-processing results on infrared images that utilized global or local information from the video and the image. In addition, in order to confirm the impact of images converted by each pre-processing technique on object detector training, we learn the YOLOX target detector for images processed by various pre-processing methods and analyze them. In particular, the results of the experiments using the CLAHE (Contrast Limited Adaptive Histogram Equalization) shows the highest detection accuracy with a mean average precision (mAP) of 81.9%.

• Smart Media Journal
• /
• v.11 no.11
• /
• pp.92-98
• /
• 2022

Identifying defects in textiles is a key procedure for quality control. This study attempted to create a model that detects defects by analyzing the images of the fabrics. The models used in the study were deep learning-based VGGNet and ResNet, and the defect detection performance of the two models was compared and evaluated. The accuracy of the VGGNet and the ResNet model was 0.859 and 0.893, respectively, which showed the higher accuracy of the ResNet. In addition, the region of attention of the model was derived by using the Grad-CAM algorithm, an eXplainable Artificial Intelligence (XAI) technique, to find out the location of the region that the deep learning model recognized as a defect in the fabric image. As a result, it was confirmed that the region recognized by the deep learning model as a defect in the fabric was actually defective even with the naked eyes. The results of this study are expected to reduce the time and cost incurred in the fabric production process by utilizing deep learning-based artificial intelligence in the defect detection of the textile industry.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.04a
• /
• pp.577-580
• /
• 2017

중합 효소 연쇄 반응 (PCR) 젤 전기영동 이미지에서 DNA 지문을 분석하기 위한 새로운 레인 검출 및 추적 알고리즘이 제안하였다. 이전에 여러 연구 결과가 보고되었지만 갑작스런 배경 밝기 차이와 구부러진 레인이 있는 이미지에서 레인을 정확하게 추출하는 것은 여전히 어려움이 있다. 우리는 평균 레인 폭과 레인 주기를 계산하기 위한 에지 기반 알고리즘을 제안한다. 본 논문에서 제안한 방법은 k-means 클러스터링 알고리즘을 이용하여 상승 에지와 하강 에지를 정확하게 추출하는 부화소(sub-pixel) 알고리즘을 적용하여 레인 폭과 주기를 추정한다. 구부러진 레인을 처리하기 위해 젤 이미지를 정상영역과 비정상영역으로 분할하고, 각 분할 된 이미지의 레인 중심을 추적한다. 우리가 제안한 방법의 성능을 평가하기 위해 534 레인을 포함한 32 개의 젤 이미지가 사용되었다. 실험 결과는 우리의 방법이 전처리 과정 없이 배경 차이와 구부러진 레인을 갖는 이미지에 강인함을 보여 주었다.