• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.1177-1179
• /
• 2003

Recently, the ground-based laser profiler is used for acquisition of 3D spatial information of a rchaeological objects. However, it is very difficult to measure complicated objects, because of a relatively low-resolution. On the other hand, texture mapping can be a solution to complement the low resolution, and to generate 3D model with higher fidelity. But, a huge cost is required for the construction of textured 3D model, because huge labor is demanded, and the work depends on editor's experiences and skills . Moreover, the accuracy of data would be lost during the editing works. In this research, using the laser profiler and a non-calibrated digital camera, a method is proposed for the automatic generation of 3D model by integrating these data. At first, region segmentation is applied to laser range data to extract geometric features of an object in the laser range data. Various information such as normal vectors of planes, distances from a sensor and a sun-direction are used in this processing. Next, an image segmentation is also applied to the digital camera images, which include the same object. Then, geometrical relations are determined by corresponding the features extracted in the laser range data and digital camera’ images. By projecting digital camera image onto the surface data reconstructed from laser range image, the 3D texture model was generated automatically.

• Journal of the Korean Society for Precision Engineering
• /
• v.4 no.2
• /
• pp.36-46
• /
• 1987

An optical technique using laser for non-contact measurement of surface roughness and form accuracy of ground surfaces is presented. It is found that, when a ground surface is illuminated by a beam of laser light, the roughness height and slope distribution has significant influence on the pattern of reflection and it maintains an unique Gaussian distribution relationship with the surface roughness. The principle idea of the optical measurement system is therefore monitor the radiation, and then calibrate it in process against surface roughness by means of necessary digital data processing. On the other hand, measuring the form accuracy of a ground surface is accomplished by using a triangular method, which is based on observing the movement of an image of a spot of light projected onto the surface. The image is focused, through a series of lenses for magnification, on a photodetector array lf line configur- ation. Then the relative movement of image and consequently the form accuracy of the surface can be obtained through appropriate calibration procedures. Experimental test showed that the optical roughness measurement technique suggested in this work is very efficient for most industrial applications being capable of monitoring the roughness heights ranging 0.1 to 0.6 .$\mu$m CLA values. And form accuracy can be measured in process with a resolution of 10 .$\mu$m.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.31 no.4
• /
• pp.391-397
• /
• 2011

Conventional film radiographic test has been generally and widely used in the inspection on the weldment for quality assurance. On the other hand, since the analog RT is well known for typical time and cost consuming method with complex process of inspection, the industry has researched various ways how to improve radiographic test technology. In this study, we verified the fact that digital RT provides a lot more benefit in effectively detecting defects, ever film details, through digital processing of image enhancement, compared to film RT. As a result, we reached conclusion that digital RT is positively able to replace the film RT in industry in part or in whole.

• Journal of the Korea Computer Graphics Society
• /
• v.6 no.4
• /
• pp.9-14
• /
• 2000

Non-Photorealistic rendering techniques, such as toon rendering, can enhance the quality of hand-drawn cell-animation images greatly with less effort. For this reason, to on rendering is one of the popular techniques used in the cell-animation image production field. The existing toon rendering techniques, however, have not been effective enough for the real-time image processing that the techniques have not been adequate for some processes that needs immediate responses such as virtual-realities, or video games. This paper will suggest the real-time toon rendering to overcome the limits through real-time outline detection and phong shading. In addition, a effective result-image is created as adding a shadow and a execution time remains by real-time through fast shadow generation algorithm.

• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.76-85
• /
• 2019

We introduce a vision-based object detection method for real-time video surveillance system in low-end edge computing environments. Recently, the accuracy of object detection has been improved due to the performance of approaches based on deep learning algorithm such as Region Convolutional Neural Network(R-CNN) which has two stage for inferencing. On the other hand, one stage detection algorithms such as single-shot detection (SSD) and you only look once (YOLO) have been developed at the expense of some accuracy and can be used for real-time systems. However, high-performance hardware such as General-Purpose computing on Graphics Processing Unit(GPGPU) is required to still achieve excellent object detection performance and speed. To address hardware requirement that is burdensome to low-end edge computing environments, We propose sub-frame analysis method for the object detection. In specific, We divide a whole image frame into smaller ones then inference them on Convolutional Neural Network (CNN) based image detection network, which is much faster than conventional network designed forfull frame image. We reduced its computationalrequirementsignificantly without losing throughput and object detection accuracy with the proposed method.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.35 no.2
• /
• pp.75-80
• /
• 2017

Recently, there has been growing interest in spatial data that combines information and communication technology with smart cities. The high-precision LiDAR (Light Dectection and Ranging) equipment is mainly used to collect three-dimensional spatial data, and the acquired data is also used to model geographic features and to manage plant construction and cultural heritages which require precision. The LiDAR equipment can collect precise data, but also has limitations because they are expensive and take long time to collect data. On the other hand, in the field of computer vision, research is being conducted on the methods of acquiring image data and performing 3D reconstruction based on image data without expensive equipment. Thus, precise 3D spatial data can be constructed efficiently by collecting and processing image data using CCTVs which are installed as infrastructure facilities in smart cities. However, this method can have an accuracy problem compared to the existing equipment. In this study, experiments were conducted and the results were analyzed to increase the number of extracted matching points by applying the feature-based method and the area-based method in order to improve the precision of 3D spatial data built with image data acquired from stereo CCTVs. For techniques to extract matching points, SIFT algorithm and PATCH algorithm were used. If precise 3D reconstruction is possible using the image data from stereo CCTVs, it will be possible to collect 3D spatial data with low-cost equipment and to collect and build data in real time because image data can be easily acquired through the Web from smart-phones and drones.

• The Journal of Korean Physical Therapy
• /
• v.21 no.1
• /
• pp.81-87
• /
• 2009

Purpose: Human brain was lateralized to dominant or non-dominant hemisphere, and could be reorganized by the processing of the motor learning. We reported four cases which showed the changes of the cortical activation patterns resulting from two weeks of training with the serial reaction time task. Methods: Four right-handed healthy subjects were recruited, who was equally divided to two training conditions (right hand training or left hand training). They were assigned to train the serial reaction time task for two weeks, which should press the corresponding four colored buttons as fast as accurately as possible when visual stimulus was presented. Before and after two weeks of training, reaction time and function magnetic resonance image (fMRI) was acquired during the performance of the same serial reaction time task as the training. Results: The reaction time was significantly decreased in all of subjects after training. Our fMRI result showed that widespread bilateral activation at the pre scanning was shifted toward the focused activation on the contralateral hemisphere with progressive motor learning. However, the bilateral activation was still remained during the performance of the non-dominant hand. Conclusion: These findings showed that the repetitive practice of the serial reaction time task led to increase the movement speed and accuracy, as described by motor learning. Such motor learning induced to change the cortical activation pattern. And, the changed pattern of the cortical activation resulting from motor learning was different each other in accordance with the hand dominance.

• Journal of the Institute of Convergence Signal Processing
• /
• v.10 no.2
• /
• pp.106-111
• /
• 2009

Real-time video stabilization technology is used in correction for the camera vibrations of the hand-held camera by hand or fixed camera by external condition. This paper is about the counterplan to cope with the vibration of the movie generated by the large external cause relatively. we use the movie stabilization parameters with the phase correlation method based the DFT to get the displacements of the current frame to the reference frame. we use the kalman filter for the efficient and stable searching works on the phase correlation map and present the proper conditions for the real-time processing through the experiments. We propose the measure to evaluate the capability of the video stabilizer which is the standard deviation of the brightness of the center block. and compare the capability for the video sequences randomly shifted and the jittered video sequences obtained from camera.

• Journal of Bio-Environment Control
• /
• v.13 no.3
• /
• pp.178-184
• /
• 2004

This study carried out to develope the vision system which automatically finds out a optimum transferring period of plants (Perilla, Platycodon grandifloums and Lactuca sativa) by using image process-ing. This system mearsured a height, long diameter and short diameter of the three plants with 20 replications. Following results were obtained on each plant. Compared with real data to be measured by hand with the vernier calipers, height, long diameter and short diameter of Perilla showed 0.5 mm average error rate with 1.7%, 4.7 mm average error rate with 3.9% and 5.5 mm average error rate with 6.9% respectively. Those of Platycodon grandifloums showed 2.4 mm with 8.1%, 3.4 mm with 7.2% and 4.0 mm with 10.4% respectively. Those of Lactuca sativa showed 4.0 mm with 9.1 %,3.4 mm with 7.2% and 3.6 mm with 9.4% respectively. The system could be used to transfer accurately the plant seedling, if the system were improved enough to reduce error rate for the optimum transferring period of a plant in the greenhouse.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.11
• /
• pp.511-520
• /
• 2016

Automatic License Plate Detection (ALPD) is a key technology for a efficient traffic control. It is used to improve work efficiency in many applications such as toll payment systems and parking and traffic management. Until recently, the hand-crafted features made for image processing are used to detect license plates in most studies. It has the advantage in speed. but can degrade the detection rate with respect to various environmental changes. In this paper, we propose a way to utilize a Faster Region based Convolutional Neural Networks (Faster R-CNN) and a Conventional Convolutional Neural Networks (CNN), which improves the computational speed and is robust against changed environments. The module based on Faster R-CNN is used to detect license plate candidate regions from images and is followed by the module based on CNN to remove False Positives from the candidates. As a result, we achieved a detection rate of 99.94% from images captured under various environments. In addition, the average operating speed is 80ms/image. We implemented a fast and robust Real-Time License Plate Detection System.