• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.205-208
• /
• 2022

Multi-exposure high dynamic range (HDR) image reconstruction, the task of reconstructing an HDR image from multiple low dynamic range (LDR) images in a dynamic scene, often produces ghosting artifacts caused by camera motion and moving objects and also cannot deal with washed-out regions due to over or under-exposures. While there has been many deep-learning-based methods with motion estimation to alleviate these problems, they still have limitations for severely moving scenes. They also require large parameter counts, especially in the case of state-of-the-art methods that employ attention modules. To address these issues, we propose a frequency domain approach based on the idea that the transform domain coefficients inherently involve the global information from whole image pixels to cope with large motions. Specifically we adopt Residual Fast Fourier Transform (RFFT) blocks, which allows for global interactions of pixels. Moreover, we also employ Depthwise Overparametrized convolution (DO-conv) blocks, a convolution in which each input channel is convolved with its own 2D kernel, for faster convergence and performance gains. We call this LFFNet (Lightweight Frequency Fusion Network), and experiments on the benchmarks show reduced ghosting artifacts and improved performance up to 0.6dB tonemapped PSNR compared to recent state-of-the-art methods. Our architecture also requires fewer parameters and converges faster in training.

• International conference on construction engineering and project management
• /
• 2020.12a
• /
• pp.313-322
• /
• 2020

The recognition of the risk hazards is a vital step to effectively prevent accidents on a construction site. The advanced development in computer vision systems and the availability of the large visual database related to construction site made it possible to take quick action in the event of human error and disaster situations that may occur during management supervision. Therefore, it is necessary to analyze the risk factors that need to be managed at the construction site and review appropriate and effective technical methods for each risk factor. This research focuses on analyzing Occupational Safety and Health Agency (OSHA) related to risk zone identification rules that can be adopted by the image recognition technology and classify their risk factors depending on the effective technical method. Therefore, this research developed a pattern-oriented classification of OSHA rules that can employ a large scale of safety hazard recognition. This research uses joint reasoning of risk zone Identification and numeric input by utilizing a stereo camera integrated with an image detection algorithm such as (YOLOv3) and Pyramid Stereo Matching Network (PSMNet). The research result identifies risk zones and raises alarm if a target object enters this zone. It also determines numerical information of a target, which recognizes the length, spacing, and angle of the target. Applying image detection joint logic algorithms might leverage the speed and accuracy of hazard detection due to merging more than one factor to prevent accidents in the job site.

• Journal of Internet Computing and Services
• /
• v.25 no.2
• /
• pp.11-19
• /
• 2024

This paper was studied abouta technology for detecting damage to temporary works equipment used in construction sites with explainable artificial intelligence (XAI). Temporary works equipment is mostly composed of steel or aluminum, and it is reused several times due to the characters of the materials in temporary works equipment. However, it sometimes causes accidents at construction sites by using low or decreased quality of temporary works equipment because the regulation and restriction of reuse in them is not strict. Currently, safety rules such as related government laws, standards, and regulations for quality control of temporary works equipment have not been established. Additionally, the inspection results were often different according to the inspector's level of training. To overcome these limitations, a method based with AI and image processing technology was developed. In addition, it was devised by applying explainableartificial intelligence (XAI) technology so that the inspector makes more exact decision with resultsin damage detect with image analysis by the XAI which is a developed AI model for analysis of temporary works equipment. In the experiments, temporary works equipment was photographed with a 4k-quality camera, and the learned artificial intelligence model was trained with 610 labelingdata, and the accuracy was tested by analyzing the image recording data of temporary works equipment. As a result, the accuracy of damage detect by the XAI was 95.0% for the training dataset, 92.0% for the validation dataset, and 90.0% for the test dataset. This was shown aboutthe reliability of the performance of the developed artificial intelligence. It was verified for usability of explainable artificial intelligence to detect damage in temporary works equipment by the experiments. However, to improve the level of commercial software, the XAI need to be trained more by real data set and the ability to detect damage has to be kept or increased when the real data set is applied.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.2
• /
• pp.95-101
• /
• 2014

This paper explains effective tracking method for a vehicle cutting in a line on the road automatically. The method employs KLT based on optical flow using play-back image sequence. Main contribution of this paper is play-back image sequence that is in order image frames for rewind direction from a reference point in time. The moment when recognizing camera can read a license plate very well can usually be the reference point in time. The biggest images of object traced can usually be obtained at this moment also. When optic flow is applied, the bigger image of the object traced can be obtained, the more feature points can be obtained. More many feature points bring good result of tracking object. After the recognizing cameras read a license plate on the vehicle suspected of cut-in-line violation, and then the system extracts the play-back image sequence from the tracking cameras for watching wide range. This paper compares using play-back image sequence as normal method for tracking to using play-forward image sequence as suggested method on the results of the experiment and also shows the suggested algorithm has a good performance that can be applied to the unmanned system for watching cut-in-line violation.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.749-754
• /
• 2011

In this study, we propose the design of optimized pRBFNNs-based face recognition system using two-dimensional Image and ASM algorithm. usually the existing 2 dimensional face recognition methods have the effects of the scale change of the image, position variation or the backgrounds of an image. In this paper, the face region information obtained from the detected face region is used for the compensation of these defects. In this paper, we use a CCD camera to obtain a picture frame directly. By using histogram equalization method, we can partially enhance the distorted image influenced by natural as well as artificial illumination. AdaBoost algorithm is used for the detection of face image between face and non-face image area. We can butt up personal profile by extracting the both face contour and shape using ASM(Active Shape Model) and then reduce dimension of image data using PCA. The proposed pRBFNNs consists of three functional modules such as the condition part, the conclusion part, and the inference part. In the condition part of fuzzy rules, input space is partitioned with Fuzzy C-Means clustering. In the conclusion part of rules, the connection weight of RBFNNs is represented as three kinds of polynomials such as constant, linear, and quadratic. The essential design parameters (including learning rate, momentum coefficient and fuzzification coefficient) of the networks are optimized by means of Differential Evolution. The proposed pRBFNNs are applied to real-time face image database and then demonstrated from viewpoint of the output performance and recognition rate.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.1
• /
• pp.9-18
• /
• 2018

Recently, the use of image stitching technology has been increasing as the number of contents based on virtual reality increases. Image Stitching is a method for matching multiple images to produce a high resolution image and a wide field of view image. The image stitching is used in various fields beyond the limitation of images generated from one camera. Image Stitching detects feature points and corresponding points to match multiple images, and calculates the homography among images using the RANSAC algorithm. Generally, corresponding points are needed for calculating conversion relation. However, the corresponding points include various types of noise that can be caused by false assumptions or errors about the conversion relationship. This noise is an obstacle to accurately predict the conversion relation. Therefore, RANSAC algorithm is used to construct an accurate conversion relationship from the outliers that interfere with the prediction of the model parameters because matching methods can usually occur incorrect correspondence points. In this paper, we propose an algorithm that extracts more accurate inliers and computes accurate transformation relations by using correspondence point relation information used in RANSAC algorithm. The correspondence point relation information uses distance ratio between corresponding points used in image matching. This paper aims to reduce the processing time while maintaining the same performance as RANSAC.

• Journal of the Korea Computer Graphics Society
• /
• v.22 no.3
• /
• pp.11-20
• /
• 2016

This paper introduces an effective method for generating 3D stereoscopic images that gives immersive 3D experiences to viewers using mobile-based binocular HMDs. Most of previous AR systems with monocular cameras have a common limitation that the same real-world images are provided to the viewer's eyes without parallax. In this paper, based on the assumption that viewers focus on the marker in the scenario of marker based AR, we recovery the binocular disparity about a camera image and a virtual object using the pose information of the marker. The basic idea is to generate the binocular disparity for real-world images and a virtual object, where the images are placed on the 2D plane in 3D defined by the pose information of the marker. For non-marker areas in the images, we apply blur effects to reduce the visual discomfort by decreasing their sharpness. Our user studies show that the proposed method for 3D stereoscopic image provides high depth feeling to viewers compared to the previous binocular AR systems. The results show that our system provides high depth feelings, high sense of reality, and visual comfort, compared to the previous binocular AR systems.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.3
• /
• pp.139-146
• /
• 2004

A 2D/3D complex optical system and its vision inspection algerian is proposed and implemented as a single probe system for high speed, precise vision inspection of the solder pastes. One pass un length labeling algorithm is proposed instead of the conventional two pass labeling algorithm for fast extraction of the 2D shape of the solder paste image from the recent line-scan camera as well as the conventional area-scan camera, and the optical probe path generation is also proposed for the efficient 2D/3D inspection. The Moire interferometry-based phase shift algerian and its optical system implementation is introduced, instead of the conventional laser slit-beam method, for the high precision 3D vision inspection. All of the time-critical algorithms are MMX SIMD parallel-coded for further speedup. The proposed system is implemented for simultaneous 2D/3D inspection of 10mm${\times}$10mm FOV with resolutions of 10 ${\mu}{\textrm}{m}$ for both x, y axis and 1 ${\mu}{\textrm}{m}$ for z axis. Experiments conducted on several nBs show that the 2D/3D inspection of an FOV, excluding an image capturing, results in high speed of about 0.011sec/0.01sec, respectively, after image capturing, with $\pm$1${\mu}{\textrm}{m}$ height accuracy.

• Journal of the Korean earth science society
• /
• v.36 no.4
• /
• pp.330-340
• /
• 2015

For this study, we developed an algorithm to estimate the total amount of clouds using sky image data from the Skyviewer equipped with CCD camera. Total cloud amount is estimated by removing mask areas of RGB (Red Green Blue) images, classifying images according to frequency distribution of GBR (Green Blue Ratio), and extracting cloud pixels from them by deciding RBR (Red Blue Ratio) threshold. Total cloud amount is also estimated by validity checks after removing sunlight area from those classified cloud pixels. In order to verify the accuracy of the algorithm that estimates total cloud amount, the research analyzed Bias, RMSE, and correlation coefficient compared to records of total cloud amount earned by human observation from the Gangwon Regional Meteorological Administration, which is in the closest vicinity of the observation site. The cases are selected four daily data from 0800 LST to 1700 LST for each season. The results of analysis showed that the Bias in total cloud amount estimated by the Skyviewer was an average of -0.8 tenth, and the RMSE was 1.6 tenths, indicating the difference in total cloud amount within 2 tenths. Also, correlation coefficient was very high, marking an average of over 0.91 in all cases, despite the distance between the two observation sites (about 4 km).

• KIPS Transactions on Software and Data Engineering
• /
• v.4 no.5
• /
• pp.225-230
• /
• 2015

In this paper, we propose an effective real-time visual loop closure detector, VILODE, which makes use of key frames and bag of visual words (BoW) based on SURF feature points. In order to determine whether the camera has re-visited one of the previously visited places, a loop closure detector has to compare an incoming new image with all previous images collected at every visited place. As the camera passes through new places or locations, the amount of images to be compared continues growing. For this reason, it is difficult for a visual loop closure detector to meet both real-time constraint and high detection accuracy. To address the problem, the proposed system adopts an effective key frame selection strategy which selects and compares only distinct meaningful ones from continuously incoming images during navigation, and so it can reduce greatly image comparisons for loop detection. Moreover, in order to improve detection accuracy and efficiency, the system represents each key frame image as a bag of visual words, and maintains indexes for them using DBoW database system. The experiments with TUM benchmark datasets demonstrates high performance of the proposed visual loop closure detector.