• Convergence Security Journal
• /
• v.19 no.5
• /
• pp.75-84
• /
• 2019

Recently, the value of video as an important data of medical information technology is increasing due to the feature of rich clinical information. On the other hand, video is also required to be de-identified as a medical image, but the existing methods are mainly specialized in the stereotyped data and still images, which makes it difficult to apply the existing methods to the video data. In this paper, we propose an automated system to index candidate elements of personal identification information on a frame basis to solve this problem. The proposed system performs indexing process using text and person detection after preprocessing by scene segmentation and color knowledge based method. The generated index information is provided as metadata according to the purpose of use. In order to verify the effectiveness of the proposed system, the indexing speed was measured using prototype implementation and real surgical video. As a result, the work speed was more than twice as fast as the playing time of the input video, and it was confirmed that the decision making was possible through the case of the production of surgical education contents.

• Journal of Biosystems Engineering
• /
• v.39 no.2
• /
• pp.142-149
• /
• 2014

Purpose: A multispectral algorithm for detection and differentiation of defective (defects on apple skin) and normal Red Delicious apples was developed from analysis of a series of hyperspectral line-scan images. Methods: A fast line-scan hyperspectral imaging system mounted on a conventional apple sorting machine was used to capture hyperspectral images of apples moving approximately 4 apples per second on a conveyor belt. The detection algorithm included an apple segmentation method and a threshold function, and was developed using three wavebands at 676 nm, 714 nm and 779 nm. The algorithm was executed on line-by-line image analysis, simulating online real-time line-scan imaging inspection during fruit processing. Results: The rapid multispectral algorithm detected over 95% of defective apples and 91% of normal apples investigated. Conclusions: The multispectral defect detection algorithm can potentially be used in commercial apple processing lines.

• Journal of Digital Convergence
• /
• v.14 no.8
• /
• pp.233-243
• /
• 2016

Due to speech recognition problems in noisy environment, Audio Visual Speech Recognition (AVSR) system, which combines speech information and visual information, has been proposed since the mid-1990s,. and lip reading have played significant role in the AVSR System. This study aims to enhance recognition rate of utterance word using only lip shape detection for efficient AVSR system. After preprocessing for lip region detection, Convolution Neural Network (CNN) techniques are applied for utterance period detection and lip shape feature vector extraction, and Hidden Markov Models (HMMs) are then used for the recognition. As a result, the utterance period detection results show 91% of success rates, which are higher performance than general threshold methods. In the lip reading recognition, while user-dependent experiment records 88.5%, user-independent experiment shows 80.2% of recognition rates, which are improved results compared to the previous studies.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.1
• /
• pp.33-45
• /
• 2011

In this paper, we present a method for extracting video objects efficiently by using the modified graph cut algorithm based on contour information. First, we extract objects at the first frame by an automatic object extraction algorithm or the user interaction. To estimate the objects' contours at the current frame, motion information of objects' contour in the previous frame is analyzed. Block-based histogram back-projection is conducted along the estimated contour point. Each color model of objects and background can be generated from back-projection images. The probabilities of links between neighboring pixels are decided by the logarithmic based distance transform map obtained from the estimated contour image. Energy of the graph is defined by predefined color models and logarithmic distance transform map. Finally, the object is extracted by minimizing the energy. Experimental results of various test images show that our algorithm works more accurately than other methods.

• Korean Journal of Agricultural Science
• /
• v.47 no.2
• /
• pp.205-217
• /
• 2020

In this study, a deep learning-based tillage boundary detection method for autonomous tillage by a tractor was developed, which consisted of image cropping, object classification, area segmentation, and boundary detection methods. Full HD (1920 × 1080) images were obtained using a RGB camera installed on the hood of a tractor and were cropped to 112 × 112 size images to generate a dataset for training the classification model. The classification model was constructed based on convolutional neural networks, and the path boundary was detected using a probability map, which was generated by the integration of softmax outputs. The results show that the F1-score of the classification was approximately 0.91, and it had a similar performance as the deep learning-based classification task in the agriculture field. The path boundary was determined with edge detection and the Hough transform, and it was compared to the actual path boundary. The average lateral error was approximately 11.4 cm, and the average angle error was approximately 8.9°. The proposed technique can perform as well as other approaches; however, it only needs low cost memory to execute the process unlike other deep learning-based approaches. It is possible that an autonomous farm robot can be easily developed with this proposed technique using a simple hardware configuration.

• Journal of Korea Society of Industrial Information Systems
• /
• v.25 no.2
• /
• pp.11-18
• /
• 2020

Recently, a series of studies on virtual try-on (VTON) using images have been published. A comparison study analyzed representative methods, SCMM-based non-deep learning method, deep learning based VITON and CP-VITON, using costumes and user images according to the posture and body type of the person, the degree of occlusion of the clothes, and the characteristics of the clothes. In this paper, we tackle the problems observed in the best performing CP-VTON. The issues tackled are the problem of segmentation of the subject, pixel generation of un-intended area, missing warped cloth mask and the cost function used in the learning, and limited the algorithm to improve it. The results show some improvement in SSIM, and significantly in subjective evaluation.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.4
• /
• pp.95-105
• /
• 2021

Road surface damage detection is essential for a comfortable driving environment and the prevention of safety accidents. Road management institutes are using automated technology-based inspection equipment and systems. As one of these automation technologies, a sensor to detect road surface damage plays an important role. For this purpose, several studies on sensors using deep learning have been conducted in recent years. Road images and label images are needed to develop such deep learning algorithms. On the other hand, considerable time and labor will be needed to secure label images. In this paper, the adversarial learning method, one of the semi-supervised learning techniques, was proposed to solve this problem. For its implementation, a lightweight deep neural network model was trained using 5,327 road images and 1,327 label images. After experimenting with 400 road images, a model with a mean intersection over a union of 80.54% and an F1 score of 77.85% was developed. Through this, a technology that can improve recognition performance by adding only road images was developed to learning without label images and is expected to be used as a technology for road surface management in the future.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.3
• /
• pp.389-395
• /
• 2022

Because of improvement of deep learning techniques, deep learning using computer vision such as classification, detection and segmentation has also been used widely at many fields. Expecially, automatic driving is one of the major fields that applies computer vision systems. Also there are a lot of works and researches to combine multiple tasks in a single network. In this study, we propose the network that predicts the individual depth of pedestrians and vehicles. Proposed model is constructed based on YOLOv3 for object detection and Monodepth for depth estimation, and it process object detection and depth estimation consequently using encoder and decoder based on hard parameter sharing. We also used attention module to improve the accuracy of both object detection and depth estimation. Depth is predicted with monocular image, and is trained using self-supervised training method.

• Journal of Korea Society of Industrial Information Systems
• /
• v.27 no.4
• /
• pp.1-10
• /
• 2022

Attempts to interpret human perspectives using computer vision have been developed in various fields. In this paper, we propose a method for evaluating the walking environment through semantic segmentation results of images from road images. First, the Kakao Map API was used to collect road images, and four-way images were collected from about 50,000 points in JeonJu. 20% of the collected images build datasets through crowdsourcing-based paired comparisons, and train various regression models using paired comparison data. In order to derive the walkability score of the image data, the ranking score is calculated using the Trueskill algorithm, which is a ranking algorithm, and the walkability and analysis using various regression models are performed using the constructed data. Through this study, it is shown that the walkability of Jeonju can be evaluated and scores can be derived through the correlation between pixel distribution classification information rather than human vision.

• Journal of Broadcast Engineering
• /
• v.27 no.5
• /
• pp.638-653
• /
• 2022

In this paper, we propose the new method of creating 3D virtual-human reflecting the pattern of clothes worn by the person in the high-resolution whole body front image and the body shape data about the person. To get the pattern of clothes, we proceed Instance Segmentation and clothes parsing using Cascade Mask R-CNN. After, we use Pix2Pix to blur the boundaries and estimate the background color and can get UV-Map of 3D clothes mesh proceeding UV-Map base warping. Also, we get the body shape data using SMPL-X and deform the original clothes and body mesh. With UV-Map of clothes and deformed clothes and body mesh, user finally can see the animation of 3D virtual-human reflecting user's appearance by rendering with the state-of-the game engine, i.e. Unreal Engine.