• Journal of the Korean Society of Marine Environment & Safety
• /
• v.26 no.6
• /
• pp.601-605
• /
• 2020

In large ships with complex structures, it is difficult to locate workers. In particular, it is not easy to detect when a worker falls down, making it difficult to respond quickly. Thus, research is being conducted to detect fallen workers using a camera or by attaching a device to the body. Existing image-based fall detection systems have been designed to detect a person's body parts; hence, it is difficult to detect them in various ships and postures. In this study, the entire fall area was extracted and deep learning was used to detect the fallen shipworker based on the image. The data necessary for learning were obtained by recording falling states at the shipyard. The amount of learning data was augmented by flipping, resizing, and rotating the image. Performance evaluation was conducted with precision, reproducibility, accuracy, and a low error rate. The larger the amount of data, the better the precision. In the future, reinforcing various data is expected to improve the effectiveness of camera-based fall detection models, and thus improve safety.

• International Journal of Computer Science & Network Security
• /
• v.23 no.5
• /
• pp.53-64
• /
• 2023

Early detection of endometrial carcinoma in uterus is essential for effective treatment. Endometrial carcinoma is the worst kind of endometrium cancer among the others since it is considerably more likely to affect the additional parts of the body if not detected and treated early. Non-invasive medical computer vision, also known as medical image processing, is becoming increasingly essential in the clinical diagnosis of various diseases. Such techniques provide a tool for automatic image processing, allowing for an accurate and timely assessment of the lesion. One of the most difficult aspects of developing an effective automatic categorization system is the absence of huge datasets. Using image processing and deep learning, this article presented an artificial endometrium cancer diagnosis system. The processes in this study include gathering a dermoscopy images from the database, preprocessing, segmentation using hybrid Fuzzy C-Means (FCM) and optimizing the weights using the Whale Optimization Algorithm (WOA). The characteristics of the damaged endometrium cells are retrieved using the feature extraction approach after the Magnetic Resonance pictures have been segmented. The collected characteristics are classified using a deep learning-based methodology called Long Short-Term Memory (LSTM) and Bi-directional LSTM classifiers. After using the publicly accessible data set, suggested classifiers obtain an accuracy of 97% and segmentation accuracy of 93%.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.313-316
• /
• 2015

Construction is one of the most dangerous job sectors, which reports tens of thousands of time-loss injuries and deaths every year. These disasters incur delays and additional costs to the projects. The safety management needs to be on the top primary tasks throughout the construction to avoid fatal accidents and to foster safe working environments. One of the safety regulations that are frequently violated is the wearing of personal protection equipment (PPE). In order to facilitate monitoring of the compliance of the PPE wearing regulations, this paper proposes a vision based method that automatically identifies whether workers wear hard hats and safety vests. The method involves three modules - human body detection, identification of safety vest wearing, and hard hat detection. First, human bodies are detected in the video frames captured by real-time on-site construction cameras. The detected human bodies are classified into with/without wearing safety vests based on the color features of their upper parts. Finally, hard hats are detected on the nearby regions of the detected human bodies and the locations of the detected hard hats and human bodies are correlated to reveal their corresponding matches. In this way, the proposed method provides any appearance of the workers without wearing hard hats or safety vests. The method has been tested on onsite videos and the results signify its potential to facilitate site safety monitoring.

• Parasites, Hosts and Diseases
• /
• v.22 no.2
• /
• pp.190-202
• /
• 1984

In an attempt to clarify the epidemiological feature of distomiasis in Tongjin riverside area, the prevalence of distomiasis in the residents and infection rates of the metacercariae in fresh-water fishes were investigated at the upper, middle and lower reaches of the river from January to April, 1984. The results obtained were summarized as follows: 1. Out of a total of 931 fresh-water fishes which composed of 33 different species, 611 fishes(65.6%) of 31 species were found positive with digenetic trematode metacercariae of 16 different species, and there were some differences in infection rates of the metacercariae among the fishes in the 3 parts of the river; 53.8% in upper, 80.7% in middle, and 61.0% in lower reaches, respectively. 2. Infection rates of the metacercariae of Exorchis oviformis, Metagonimus yokogawai, Echinochasmus japenicus, Metorchis orientalis and Clonorchis sinensis in the fishes were 48%,29%, 115, 7.9% and 6.3oA, respectively. 3. The average number of the encysted larvae of Clonorchis found in fish body/gram showed 4.44 in Pseudorasbera larva, Gnathepegon coreanus (1.2), Microphysogoio yaluensis (0.76), Abbottina springeri (0.4), Acanthorhodeus asmussi (0.21) and Cultriculus eigenmanni (0.17), respectively. 4. The average number of the metacercariae of Metagonimus found in fish body/gram disclosed 34.01 in Zacco platypus, Zacco temmincki (16.46), Carassius carassius (5.35), Moroco oxycephalus (1.54) , Aphyocypris chinensis (1.5) and etc., respectively. 5. Detection rates of the eggs of Clonorchis and Metagonimus among residents were 1.1% and 0.8%, respectively, out of a total 923 Persons.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.26 no.6
• /
• pp.440-446
• /
• 2006

The reactor vessel body and closure head are fastened with the stud bolt that is one of crucial parts for safety of the reactor vessels in nuclear power plants. It is reported that the stud bolt is often experienced by fatigue cracks initiated at threads. Stud bolts are inspected by the ultrasonic technique during the overhaul periodically for the prevention of failure which leads to radioactive leakage from the nuclear reactor. The conventional ultrasonic inspection for stud bolts was mainly conducted by reflected echo method based on shadow effect. However, in this technique, there were numerous spurious signals reflected from every oblique surfaces of the thread. In this study, ultrasonic phased array technique was applied to investigate detectability of flaws in stud bolts and characteristics of ultrasonic images corresponding to different scanning methods, that is, sector and linear scan. For this purpose, simplified stud bolt specimens with artificial defects of various depths were prepared.

• Proceedings of the KSMRM Conference
• /
• 2003.10a
• /
• pp.100-100
• /
• 2003

In brain MR imaging, contrast-enhanced study is important in the detection and characterization of lesions. As a postcontrast brain MR imaging, conventional T1 weighted imaging has been usually used. Magnetization transfer imaging has been used to increase conspicuity of enhancing lesions. In addition, fat-suppression imaging can be used as in other parts of the body. Recently, FLAIR sequence has been reported to be useful in detecting subarachnoid, meningeal, and subdural abnormalities. In this exhibit, we demonstrate basic principles and typical appearances of various pulse sequences that can be used as a postcontrast brain MR imaging in children. Furthermore, we discuss imaging strategies to increase clinical usefulness of postcontrast brain MR imaging for specific abnormalities. The advantages and disadvantages of each pulse sequence are also discussed.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2011.11a
• /
• pp.388-390
• /
• 2011

인터넷에서 정보의 쉬운 접근성으로 청소년들에게 무방비로 노출되어 있는 음란물을 자동으로 제어하는 연구는 다양하게 진행되고 있다. 본 연구는 음란 이미지를 자동으로 판단하는 방법에 대한 것으로, 특히 좌우로 누워있는 음란 이미지를 감지하는 방법을 제안하였다. 제안된 알고리즘의 유용성 검증을 위해 실험을 통해 분석하였다. 실험결과는 만족스러운 성능을 보여주지 않았고 몇 가지 추가적인 문제도 도출 되었다.

• Journal of KIISE
• /
• v.41 no.11
• /
• pp.921-926
• /
• 2014

Face detection refers to the process extracting facial regions in an input image, which can improve speed and accuracy of recognition or authorization system, and has diverse applicability. Since conventional works have tried to detect faces based on the whole shape of faces, its detection performance can be degraded by occlusion made with accessories or parts of body. In this paper we propose a method combining local feature descriptors and probability modeling in order to detect partially occluded face effectively. In training stage, we represent an image as a set of local feature descriptors and estimate a statistical model for normal faces. When the test image is given, we find a region that is most similar to face using our face model constructed in training stage. According to experimental results with benchmark data set, we confirmed the effect of proposed method on detecting partially occluded face.

• Resources Recycling
• /
• v.25 no.4
• /
• pp.11-22
• /
• 2016

In global automobile industry, the remanufacturing for used products has the merit to be reduced nearly 80 percent of energy consumption and resources of new product. The objective of this paper is the analysis of detailed remanufacturing processes about research object and failure modes of each process of throttle body which is one of automobile parts, to draw a FMEA and determine the degree of seriousness (S), detection (D) and occurrence (O) of many failures. And we compared the current RPN method of being used to calculate values of RPN with three suggested methods. : Summation method, Square root method, Volume method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1202-1205
• /
• 2005

This paper describes a method for vision-based person identification that can detect, track, and recognize person from video using multiple cues: height and dressing colors. The method does not require constrained target's pose or fully frontal face image to identify the person. First, the system, which is connected to a pan-tilt-zoom camera, detects target using motion detection and human cardboard model. The system keeps tracking the moving target while it is trying to identify whether it is a human and identify who it is among the registered persons in the database. To segment the moving target from the background scene, we employ a version of background subtraction technique and some spatial filtering. Once the target is segmented, we then align the target with the generic human cardboard model to verify whether the detected target is a human. If the target is identified as a human, the card board model is also used to segment the body parts to obtain some salient features such as head, torso, and legs. The whole body silhouette is also analyzed to obtain the target's shape information such as height and slimness. We then use these multiple cues (at present, we uses shirt color, trousers color, and body height) to recognize the target using a supervised self-organization process. We preliminary tested the system on a set of 5 subjects with multiple clothes. The recognition rate is 100% if the person is wearing the clothes that were learned before. In case a person wears new dresses the system fail to identify. This means height is not enough to classify persons. We plan to extend the work by adding more cues such as skin color, and face recognition by utilizing the zoom capability of the camera to obtain high resolution view of face; then, evaluate the system with more subjects.