• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.3
• /
• pp.271-278
• /
• 2014

In spite of that stop line and crosswalk should be aware of the most basic objects in transportation system, its features extracted are very limited. In addition to image-based recognition technology, laser and RF, GPS/INS recognition technology, it is difficult to recognize. For this reason, the limited research in this area has been done. In this paper, the algorithm to recognize the stop line and crosswalk is designed and implemented using image-based recognition technology with the images input through a vision sensor. This algorithm consists of three functions.; One is to select the area, in advance, needed for feature extraction in order to speed up the data processing, 'Region of Interest', another is to process the images only that white color is detected more than a certain proportion in order to remove the unnecessary operation, 'Color Pattern Inspection', the other is 'Feature Extraction and Recognition', which is to extract the edge features and compare this to the previously-modeled one to identify the stop line and crosswalk. For this, especially by using case based feature comparison algorithm, it can identify either both stop line and crosswalk exist or just one exists. Also the proposed algorithm is to develop existing researches by comparing and analysing effect of in-vehicle camera installation and changes in recognition rate of distance estimation and various constraints such as backlight and shadow.

• Journal of Radiation Protection and Research
• /
• v.42 no.4
• /
• pp.197-204
• /
• 2017

Background: X-ray imaging detectors for the nondestructive cargo container inspection using MeV-energy X-rays should accurately portray the internal structure of the irradiated container. Internal and external factors can cause noise, affecting image quality, and scattered radiation is the greatest source of noise. To obtain a high-performance transmission image, the influence of scattered radiation must be minimized, and this can be accomplished through several methods. The scatter rejection method using an anti-scatter grid is the preferred method to reduce the impact of scattered radiation. In this paper, we present an evaluation the characteristics of the signal and noise according to physical and material changes in the anti-scatter grid of the imaging detector used in cargo container scanners. Materials and Methods: We evaluated the characteristics of the signal and noise according to changes in the grid ratio and the material of the anti-scatter grid in an X-ray image detector using MCNP6. The grid was composed of iron, lead, or tungsten, and the grid ratio was set to 2.5, 12.5, 25, or 37.5. X-ray spectrum sources for simulation were generated by 6- and 9-MeV electron impacts on the tungsten target using MCNP6. The object in the simulation was designed using metallic material of various thicknesses inside the steel container. Using the results of the computational simulation, we calculated the change in the scatter-to-primary ratio and the signal-to-noise ratio improvement factor according to the grid ratio and the grid material, respectively. Results and Discussion: Changing the grid ratios of the anti-scatter grid and the grid material decreased the scatter linearly, affecting the signal-to-noise ratio. Conclusion: The grid ratio and material of the anti-scatter grid affected the response characteristics of a container scanner using high-energy X-rays, but to a minimal extent; thus, it may not be practically effective to incorporate anti-scatter grids into container scanners.

• Journal of Intelligence and Information Systems
• /
• v.15 no.2
• /
• pp.33-48
• /
• 2009

Capsule endoscopy is one of the most remarkable inventions in last ten years. Causing less pain for patients, diagnosis for entire digestive system has been considered as a most convenience method over a normal endoscope. However, it is known that the diagnosis process typically requires very long inspection time for clinical experts because of considerably many duplicate images of same areas in human digestive system due to uncontrollable movement of a capsule endoscope. In this paper, we propose a method for clinical diagnosticians to get highly valuable information from capsule-endoscopy video. Our software system consists of three global maps, such as movement map, characteristic map, and brightness map, in temporal domain for entire sequence of the input video. The movement map can be used for effectively removing duplicated adjacent images. The characteristic and brightness maps provide frame content analyses that can be quickly used for segmenting regions or locating some features(such as blood) in the stream. Our experiments show the results of four patients having different health conditions. The result maps clearly capture the movements and characteristics from the image frames. Our method may help the diagnosticians quickly search the locations of lesion, bleeding, or some other interesting areas.

• Journal of Korean Neurosurgical Society
• /
• v.62 no.4
• /
• pp.376-381
• /
• 2019

Objective : To define optimal method that calculate the safe direction of cervical pedicle screw placement using computed tomography (CT) image based three dimensional (3D) cortical shell model of human cervical spine. Methods : Cortical shell model of cervical spine from C3 to C6 was made after segmentation of in vivo CT image data of 44 volunteers. Three dimensional Cartesian coordinate of all points constituting surface of whole vertebra, bilateral pedicle and posterior wall were acquired. The ideal trajectory of pedicle screw insertion was defined as viewing direction at which the inner area of pedicle become largest when we see through the biconcave tubular pedicle. The ideal trajectory of 352 pedicles (eight pedicles for each of 44 subjects) were calculated using custom made program and were changed from global coordinate to local coordinate according to the three dimensional position of posterior wall of each vertebral body. The transverse and sagittal angle of trajectory were defined as the angle between ideal trajectory line and perpendicular line of posterior wall in the horizontal and sagittal plane. The averages and standard deviations of all measurements were calculated. Results : The average transverse angles were $50.60^{\circ}{\pm}6.22^{\circ}$ at C3, $51.42^{\circ}{\pm}7.44^{\circ}$ at C4, $47.79^{\circ}{\pm}7.61^{\circ}$ at C5, and $41.24^{\circ}{\pm}7.76^{\circ}$ at C6. The transverse angle becomes more steep from C3 to C6. The mean sagittal angles were $9.72^{\circ}{\pm}6.73^{\circ}$ downward at C3, $5.09^{\circ}{\pm}6.39^{\circ}$ downward at C4, $0.08^{\circ}{\pm}6.06^{\circ}$ downward at C5, and $1.67^{\circ}{\pm}6.06^{\circ}$ upward at C6. The sagittal angle changes from caudad to cephalad from C3 to C6. Conclusion : The absolute values of transverse and sagittal angle in our study were not same but the trend of changes were similar to previous studies. Because we know 3D address of all points constituting cortical shell of cervical vertebrae. we can easily reconstruct 3D model and manage it freely using computer program. More creative measurement of morphological characteristics could be carried out than direct inspection of raw bone. Furthermore this concept of measurement could be used for the computing program of automated robotic screw insertion.

• Journal of the Korean Recycled Construction Resources Institute
• /
• v.9 no.3
• /
• pp.303-310
• /
• 2021

Cracks in bridges are important factors that indicate the condition of bridges and should be monitored periodically. However, a visual inspection conducted by a human expert has problems in cost, time, and reliability. Therefore, in recent years, researches to apply a deep learning model are started to be conducted. Deep learning requires sufficient data on the situations to be predicted, but bridge crack data is relatively difficult to obtain. In particular, it is difficult to collect a large amount of crack data in a specific situation because the shape of bridge cracks may vary depending on the bridge's design, location, and construction method. This study developed a crack detection model that generates and trains insufficient crack data through a Generative Adversarial Network. GAN successfully generated data statistically similar to the given crack data, and accordingly, crack detection was possible with about 3% higher accuracy when using the generated image than when the generated image was not used. This approach is expected to effectively improve the performance of the detection model as it is applied when crack detection on bridges is required, though there is not enough data, also when there is relatively little or much data f or one class.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.23 no.6
• /
• pp.549-558
• /
• 2021

In order to keep infrastructures such as tunnels and underground facilities safe, cracks of concrete lining in tunnel should be detected by regular inspections. Since regular inspections are accomplished through manual efforts using maintenance lift vehicles, it brings about traffic jam, exposes works to dangerous circumstances, and deteriorates consistency of crack inspection data. This study aims to provide methodology to automatically extract cracks from tunnel concrete lining images generated by the existing tunnel image acquisition system. Specifically, we train a deep learning based semantic segmentation model with open dataset, and evaluate its performance with the dataset from the existing tunnel image acquisition system. In particular, we compare the model performance in case of using all of a public dataset, subset of the public dataset which are related to tunnel surfaces, and the tunnel-related subset with negative examples. As a result, the model trained using the tunnel-related subset with negative examples reached the best performance. In the future, we expect that this research can be used for planning efficient model training strategy for crack detection.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.6
• /
• pp.409-418
• /
• 2021

As the amount of construction for aerial photography increases, the need for automation of quality inspection is emerging. In this study, an experiment was performed to classify or detect clouds in aerial photos using deep learning techniques. Also, classification and detection were performed by including satellite images in the learning data. As algorithms used in the experiment, GoogLeNet, VGG16, Faster R-CNN and YOLOv3 were applied and the results were compared. In addition, considering the practical limitations of securing erroneous images including clouds in aerial images, we also analyzed whether additional learning of satellite images affects classification and detection accuracy in comparison a training dataset that only contains aerial images. As results, the GoogLeNet and YOLOv3 algorithms showed relatively superior accuracy in cloud classification and detection of aerial images, respectively. GoogLeNet showed producer's accuracy of 83.8% for cloud and YOLOv3 showed producer's accuracy of 84.0% for cloud. And, the addition of satellite image learning data showed that it can be applied as an alternative when there is a lack of aerial image data.

• Journal of IKEEE
• /
• v.25 no.3
• /
• pp.493-500
• /
• 2021

In this paper, we propose the development of deep learning structure to improve quality of polygonal containers. The deep learning structure consists of a convolution layer, a bottleneck layer, a fully connect layer, and a softmax layer. The convolution layer is a layer that obtains a feature image by performing a convolution 3x3 operation on the input image or the feature image of the previous layer with several feature filters. The bottleneck layer selects only the optimal features among the features on the feature image extracted through the convolution layer, reduces the channel to a convolution 1x1 ReLU, and performs a convolution 3x3 ReLU. The global average pooling operation performed after going through the bottleneck layer reduces the size of the feature image by selecting only the optimal features among the features of the feature image extracted through the convolution layer. The fully connect layer outputs the output data through 6 fully connect layers. The softmax layer multiplies and multiplies the value between the value of the input layer node and the target node to be calculated, and converts it into a value between 0 and 1 through an activation function. After the learning is completed, the recognition process classifies non-circular glass bottles by performing image acquisition using a camera, measuring position detection, and non-circular glass bottle classification using deep learning as in the learning process. In order to evaluate the performance of the deep learning structure to improve quality of polygonal containers, as a result of an experiment at an authorized testing institute, it was calculated to be at the same level as the world's highest level with 99% good/defective discrimination accuracy. Inspection time averaged 1.7 seconds, which was calculated within the operating time standards of production processes using non-circular machine vision systems. Therefore, the effectiveness of the performance of the deep learning structure to improve quality of polygonal containers proposed in this paper was proven.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.2
• /
• pp.786-792
• /
• 2013

In this study was explored minimize side effects due to the additional injection of contrast medium and maintaining a high resolution imaging applied to the inspection and analysis of the contrast medium that affect the peak time biomechanics factors. Included 48 patients using the test bolus method, after measuring a patient's biomechanics factors of inspection before and during the test, correlation between contrast medium peak time and learn, matches the regression equation calculated and measured contrast medium peak time was assessed by the Bland Altman plot. Research result, inspections of SBP, HR contrast medium peak time and a significant negative correlation was, step 1, every increase, the contrast medium peak time significantly to -0.018 and -0.159 decreased, a fairly high concordance no difference between the two method. In conclusion, the regression equation using the existing methods, while maintaining excellent image quality that contrast medium is reduced to a patient, it can conclude that the alternative to the existing methods.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.24 no.3
• /
• pp.275-281
• /
• 2004

Socket weldment used to change the flow direction of fluid nay have flaws such as lack of fusion and cracks. Liquid penetrant testing or Radiography testing have been applied as NDT methods for flaw detection of the socket weldment. But it is difficult to detect the flaw inside of the socket weldment with these methods. In order to inspect the flaws inside the socket weldment, a ultrasonic testing method is established and a ultrasonic transducer and automated ultrasonic testing system are developed for the inspection. The automated ultrasonic testing system is based on the portable personal computer and operated by the program based Windows 98 or 2000. The system has a pulser/receiver, 100MHz high speed A/D board, and basic functions of ultrasonic flaw detector using the program. For the automated testing, motion controller board of ISA interface type is developed to control the 4-axis scanner and a real time iC-scan image of the automated testing is displayed on the monitor. A flaws with the size of less than 1mm in depth are evaluated smaller than its actual site in the testing, but the flaws larger than 1mm appear larger than its actual size on the contrary. This tendency is shown to be increasing as the flaw size increases. h reliable and objective testing results are obtained with the developed system, so that it is expected that it can contribute to safety management and detection of repair position of pipe lines of nuclear power plants and chemical plants.