• Journal of Animal Science and Technology
• /
• v.65 no.6
• /
• pp.1254-1269
• /
• 2023

This study aims to predict the change in corn share according to the grazing of 20 gestational sows in a mature corn field by taking images with a camera-equipped unmanned air vehicle (UAV). Deep learning based on convolutional neural networks (CNNs) has been verified for its performance in various areas. It has also demonstrated high recognition accuracy and detection time in agricultural applications such as pest and disease diagnosis and prediction. A large amount of data is required to train CNNs effectively. Still, since UAVs capture only a limited number of images, we propose a data augmentation method that can effectively increase data. And most occupancy prediction predicts occupancy by designing a CNN-based object detector for an image and counting the number of recognized objects or calculating the number of pixels occupied by an object. These methods require complex occupancy rate calculations; the accuracy depends on whether the object features of interest are visible in the image. However, in this study, CNN is not approached as a corn object detection and classification problem but as a function approximation and regression problem so that the occupancy rate of corn objects in an image can be represented as the CNN output. The proposed method effectively estimates occupancy for a limited number of cornfield photos, shows excellent prediction accuracy, and confirms the potential and scalability of deep learning.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.411-413
• /
• 2002

In this paper we researched about the ultrasonic cane which aids the blind to walk. We used ultrasonic in recognizing the object and implemented pulse counting method in measuring the distance. The distance measuring system consists of transmitter unit, receiver unit and micro-processor. We used broadband ultrasonic sensors in transmitter unit and receiver unit. The blind is able to recognize the distance between obstacles and himself as it used a vibration system.

• Journal of the Society of Naval Architects of Korea
• /
• v.54 no.2
• /
• pp.143-150
• /
• 2017

Previous methods to evaluate the risk of dropped objects rely on personnel experience of the engineer or operator without analyzed data. However analyzing historical statistic data is the best approach to find the safest operation route and to achieve more reasonable and reliable calculation results. By counting the failure frequency and fatal accident rate the risk can be quantified, and so controlled or mitigated with best economical risk reducing measures. This analysis gives a crane operator with useful information for selecting the best crane operation route, and a designer with an estimation of risk level for the dropped objects from a safety point of view.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.6
• /
• pp.191-201
• /
• 2019

A fast and accurate integrated traffic object detection framework was proposed and developed, harnessing a computer-vision based deep-learning approach performing automatic object detections, a multi object tracking technology, and video pre-processing tools. The proposed method is capable of detecting traffic object such as autos, buses, trucks and vans from video recordings taken under a various kinds of external conditions such as stability of video, weather conditions, video angles, and counting the objects by tracking them on a real-time basis. By creating plausible experimental scenarios dealing with various conditions that likely affect video quality, it is discovered that the proposed method achieves outstanding performances except for the cases of rain and snow, thereby resulting in 98% ~ 100% of accuracy.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.3
• /
• pp.551-562
• /
• 2024

This study relates to a method and device for extracting and tracking moving objects. In particular, objects are extracted using different images between adjacent images, and the location information of the extracted object is continuously transmitted to provide accurate location information of at least one moving object. It relates to a method and device for extracting and tracking moving objects based on tracking moving objects. People tracking, which started as an expression of the interaction between people and computers, is used in many application fields such as robot learning, object counting, and surveillance systems. In particular, in the field of security systems, cameras are used to recognize and track people to automatically detect illegal activities. The importance of developing a surveillance system, that can detect, is increasing day by day.

• Journal of Mechanical Science and Technology
• /
• v.15 no.10
• /
• pp.1369-1379
• /
• 2001

This paper focuses on a development of a sensor system measuring locations of a vehicle to localize a mobile robot while it tracks on the track (location sensor) . Also it focuses on a system configuration identifying the vehicle's orientation and distance from the object while it is stationary at certain station (position sensor) . As for the location sensor it consists of a set of sensors with a combined guiding and counting sensor, and an address-coded sensor to localize the vehicle while moving on the rail. For the position sensor a PSD (Position Sensitive Device) sensor with photo-switches sensor to measure the offset and orientation of the vehicle at each station is introduced. Both sensor systems are integrated with a microprocessor as a data relay to the main computer controlling the vehicle. The location sensor system is developed and its performance for a mobile robot is verified by experiments. The position measuring system is proposed and is robust to the environmental variation. Moreover, the two kinds of sensor systems guarantee a low cost application and high reliability.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39B no.7
• /
• pp.459-466
• /
• 2014

In this paper, we suggest an efficient image processing method for fast paper currency counting with pattern recognition. The patterns are consisted of feature data in each note object extracted from full reflection image of notes and a general contact image sensor(CIS) is used to aggregate the feature images. The proposed pattern recognition algorithm can endure image variation when the paper currency is scanned because it is not sensitive to changes of image resulting in successful note recognition. We tested 100 notes per denomination and currency of several countries including Korea, U.S., China, EU, Britain and Turkey. To ensure the reliability of the result, we tested a total of 10 times per each direction of notes. We can conclude that this algorithm will be applicable to commercial product because of its successful recognition rates. The 100% recognition rates are obtained in almost cases with exceptional case of 99.9% in Euro and 99.8% in Turkish Lira.

• Journal of Radiation Protection and Research
• /
• v.42 no.2
• /
• pp.83-90
• /
• 2017

Background: Different cases exist in the measurement of thyroid radiobioassays owing to the individual characteristics of the subjects, especially the potential variation in the counting efficiency. An In situ Object Counting System (ISOCS) was developed to perform an efficiency calibration based on the Monte Carlo calculation, as an alternative to conventional calibration methods. The purpose of this study is to evaluate the applicability of ISOCS to thyroid radiobioassays by comparison with a conventional thyroid monitoring system. Materials and Methods: The efficiency calibration of a portable high-purity germanium (HPGe) detector was performed using ISOCS software. In contrast, the conventional efficiency calibration, which needed a radioactive material, was applied to a scintillator-based thyroid monitor. Four radioiodine samples that contained $^{125}I$ and $^{131}I$ in both aqueous solution and gel forms were measured to evaluate radioactivity in the thyroid. ANSI/HPS N13.30 performance criteria, which included the relative bias, relative precision, and root-mean-squared error, were applied to evaluate the performance of the measurement system. Results and Discussion: The portable HPGe detector could measure both radioiodines with ISOCS but the thyroid monitor could not measure $^{125}I$ because of the limited energy resolution of the NaI(Tl) scintillator. The $^{131}I$ results from both detectors agreed to within 5% with the certified results. Moreover, the $^{125}I$ results from the portable HPGe detector agreed to within 10% with the certified results. All measurement results complied with the ANSI/HPS N13.30 performance criteria. Conclusion: The results of the intercomparison program indicated the feasibility of applying ISOCS software to direct thyroid radiobioassays. The portable HPGe detector with ISOCS software can provide the convenience of efficiency calibration and higher energy resolution for identifying photopeaks, compared with a conventional thyroid monitor with a NaI(Tl) scintillator. The application of ISOCS software in a radiation emergency can improve the response in terms of internal contamination monitoring.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1142-1146
• /
• 2004

The ultra precision linear stage is an essential device in the fields of MEMS and Bio technology. A piezo electric motor is widely used for its better linear characteristics, faster response time, and smaller size than conventional electro-magnetic actuator. We develop a new inchworm type motor to implement an actuator-integrated a long stroke linear stage which can move fast. To implement a servo system, we use a heterodyne interferometer as a position sensor, and we propose a new measurement technique using I/Q demodulator, and we propose a counting method to measure the position of fast moving object with low cost circuitry. The characteristics of the actuator and servo system are evaluated by measuring its displacement with a commercial laser interferometer.

• Journal of Satellite, Information and Communications
• /
• v.8 no.4
• /
• pp.26-29
• /
• 2013

In order to estimate people crowdedness in public area, the texture based method or motion based method can be used. In this paper we have proposed a mixed method. By designating the region of interest, we made the degree of crowdedness more accurate. The feature normalization also reduced the image distortion which results from difference of camera angle. The proposed system was optimized to real time embedded hardware system.