• The Journal of Korea Robotics Society
• /
• v.6 no.1
• /
• pp.18-26
• /
• 2011

This paper presents a method, Exposure Controlled Temporal Filtering (ECF), applied to visual motion tracking, that can cancel the temporal aliasing of periodic vibrations of cameras and fluctuations in illumination through the control of exposure time. We first present a theoretical analysis of the exposure induced image time integration process and how it samples sensor impingent light that is periodically fluctuating. Based on this analysis we develop a simple method to cancel high frequency vibrations that are temporally aliased onto sampled image sequences and thus to subsequent motion tracking measurements. Simulations and experiments using the 'Center of Gravity' and Normalized Cross-Correlation motion tracking methods were performed on a microscopic motion tracking system to validate the analytical predictions.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.15 no.2
• /
• pp.245-252
• /
• 1997

The development of digital camera and advancement of computer processor could simplify the acquisition and the analysis of digial image, and be the real-time processing by the digital photogrammetry. This study is about to enhancement of the image measuring accuracy by the centroid methods. We were able to determine more effective centroid measuring methods and suitable target shape as the development of analysis system and actualize semi-automatic measuring of digital image. And we can supply the weakness of non-metric camera for the geometric internal accuracy of digital image as the correct of Kodak DCS200 camera 8008s lens distortion.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.2
• /
• pp.907-923
• /
• 2020

For the severe haze situation in the Beijing-Tianjin-Hebei region, conventional fine particulate matter (PM2.5) concentration prediction methods based on pollutant data face problems such as incomplete data, which may lead to poor prediction performance. Therefore, this paper proposes a method of predicting the PM2.5 concentration based on image analysis technology that combines image data, which can reflect the original weather conditions, with currently popular machine learning methods. First, based on local parameter estimation, autoregressive (AR) model analysis and local estimation of the increase in image blur, we extract features from the weather images using an approach inspired by free energy and a no-reference robust metric model. Next, we compare the coefficient energy and contrast difference of each pixel in the AR model and then use the percentages to calculate the image sharpness to derive the overall mass fraction. Furthermore, the results are compared. The relationship between residual value and PM2.5 concentration is fitted by generalized Gauss distribution (GGD) model. Finally, nonlinear mapping is performed via the wavelet neural network (WNN) method to obtain the PM2.5 concentration. Experimental results obtained on real data show that the proposed method offers an improved prediction accuracy and lower root mean square error (RMSE).

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.112-115
• /
• 2022

The process of deep learning usually needs to deal with massive data which has greatly limited the development of deep learning technologies today. Convolutional Neural Network (CNN) structure is often used to solve image classification problems. However, a large number of images may be required in order to train an image in CNN, which is a heavy burden for existing computer systems to handle. If the image data can be compressed under the premise that the computer hardware system remains unchanged, it is possible to train more datasets in deep learning. However, image compression usually adopts the form of lossy compression, which will lose part of the image information. If the lost information is key information, it may affect learning performance. In this paper, we will analyze the effect of image compression on deep learning performance on CNN-based cat and dog classification. Through the experiment results, we conclude that the compression of images does not have a significant impact on the accuracy of deep learning.

• Journal of radiological science and technology
• /
• v.26 no.3
• /
• pp.7-11
• /
• 2003

Currently, many hospitals are hastening to introduce digital radiography systems. This is a direct result of the intentions to improve medical services and to digitalize radiology information systems, and is also leading to the improvement of medical imaging technology. Throughout F/S system's long history, many people have researched the image quality and dosage concerning these systems, and as a result, huge improvements in the dosage of patients were possible. Similarly, I believe that DR systems need the same kind of effort. Of course, decreases in dosage that ignore image quality are unthinkable. The results of experiments conducted by five hospitals during a period of 3 months brought to us the conclusions listed below. 1. Based on the comparison and analysis of the exposure control of F/S systems and DR systems, DR systems generally showed higher exposure control for parts of the phantom that became thicker, and the exposure control improved rapidly as the thickness increased. 2. DR systems still proved to be somewhat deficient in resolution measurements compared to existing F/S systems. The image processing part of DR systems contributed much to these result. 3. Under conditions used clinically, the dosage measurements of DR systems were generally higher regardless of region. 4. According to the evaluation of image quality, DR systems showed a higher degree of satisfaction as the thickness of the region became thinner. As mentioned above and based on the mutual relationship experiments between the dosage and image quality of F/S systems and DR systems, research to increase the satisfaction of DR systems must be considered.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.05a
• /
• pp.834-835
• /
• 2016

In recent years, advances in technology, IT is rapidly growing. Accordingly, real time image processing and multiple platforms, providing compatibility with OpenCV for image processing technology research on actively in progress. At present, different, comparing the images to determine the similarity is low, the system will match the rate of people using the analogue figures to determine the system is for the most part. In this paper, Template Matching of OpenCV and Feature Matching utilizing different images to determine the similarity between digital values for the system. A comparison of the features of a specific point on the screen the image to extract the same feature in a different size, you can compare the features of the target image recognized as compared to three historic castle in comparison, verification. This is the voice and image recognition and analysis, check the matching rate readings than in Zhengzhou treatment techniques are available. The future of forensic and other image processing technologies for OpenCV studies will be needed to feed.

• The KIPS Transactions:PartA
• /
• v.10A no.4
• /
• pp.397-404
• /
• 2003

Dedicated-purpose embedded Java system usually takes such model that all class files are converted into a single ROM Image by the ROMizer in the host computer, and then the Java virtual machine in the embedded system executes the image. Defining the ROM Image is a very important issue for embedded system with limited memory resource and low-performance processor since the format directly influences on the memory usage and effectiveness of accessing entries in classes. In this paper we have analyzed the ROMizer and especially the format of the ROM image implemented in the simpleRTJ embedded Jana virtual machine. The analysis says that memory space can be saved up to 50% compared to the original class file and access speed exceeds up to six times with the use of the ROMizer. The result of this study will be applied to develop a more efficient ROMizer for a ROM-based embedded Java system.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.2
• /
• pp.270-275
• /
• 2012

This paper presents a portable and compact optical device which can conveniently be used to perform a functional analysis of human liver function. The proposed system employed red/green LEDs, as a light source, and CMOS image sensor, which is commonly used in cellular phones. With this system, several blood serum samples have been evaluated for liver functional analysis by measuring light absorption level through the blood serum samples depending on aspartate aminotransferase (AST), alanine aminotransferase (ALT), and total bilirubin concentration. The light absorption through the blood serum samples containing AST, ALT, or total bilirubin can provide their concentrations. The green light absorption is more sensitive to the concentration of AST or ALT, and the red light absorption is more sensitive to the total bilirubuin concentration. Additional calibration steps were performed by using a MATLAB program in order to eliminate the light scattering effects from the extraneous particles existing in each blood serum sample. From the blind test, three standard light intensity curves through each enzyme have been obtained and the enzyme concentration values have been compared to those obtained from a commercially available biochemistry analyzer (Toshiba 200 FR). The average percent difference in the obtained concentrations from two systems for AST, ALT, and total bilirubin concentration came out to be 7.79%, 7.98%. and 7.56%, respectively, with the adjusted coefficient of determination (R2) higher than 0.98. This system can possibly lead to a low-cost and simple system that can be used as a point-of-care (POC) system in a condition without advanced equipments.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 1996.06c
• /
• pp.834-841
• /
• 1996

Egg production in Korea is becoming automated with a large scale farm. Although many operations in egg production have been and cracks are regraded as a critical problem. A computer vision system was built to generate images of a single , stationary egg. This system includes a CCD camera, a frame grabber board, a personal computer (IBM PC AT 486) and an incandescent back lighting system. Image processing algorithms were developed to inspect egg shell and to sort eggs. Those values of both gray level and area of dark spots in the egg image were used as criteria to detect holes in egg and those values of both area and roundness of dark spots in the egg and those values of both area and roundness of dark spots in the egg image were used to detect cracks in egg. Fro a sample of 300 eggs. this system was able to correctly analyze an egg for the presence of a defect 97.5% of the time. The weights of eggs were found to be linear to both the projected area and the perimeter of eggs v ewed from above. Those two values were used as criteria to sort eggs. Accuracy in grading was found to be 96.7% as compared with results from weight by electronic scale.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.2
• /
• pp.321-327
• /
• 2008

In recently, researches which appling satellite image are introduced target to the forest area. Especially, it is satellite image analysis's advantage that collecting information of terrain at the direct accesses are dangerous and impassible area. But, the studies approaching to a inflectional paradigm of forests and change detection system for the distinguished forests type are leaves much to be desired. In this study, therefore, change of forests type was analyzed using Landsat TM satellite image which have multi-spectral bands. Furthermore, change detection system for forests type was constructed on web for the periodical monitoring.