• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.1
• /
• pp.1-11
• /
• 2001

This paper presents a learning-based approach for the construction of license Plate recognition system. The system consist of three modules. They are respectively, car detection module, license plate recognition module and recognition module. Car detection module detects a car in the given image sequence obtained from the camera with simple color-based approach. Segmentation module extracts the license plate in detect car image using neural network as filters for analyzing the color and texture properties of license plate. Recognition module then reads characters in detected license plate with support vector machine (SVM)-based characters recognizer. The system has been tested from parking lot and tollgate, etc. and have show the following performances on average: Car detect rate 100%, segmentation rate 97.5%, and character recognition rate about 97.2%. Overall system performances is 94.7% and processing time is one sec. Then our propose system does well using real world.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.7 no.5
• /
• pp.33-52
• /
• 2008

This research intended to study a vehicle detection method for traffic volume, speed, stoping, parking activity, etc. using real time image processing and to propose the more accurate, environment adaptive image processing algorithm. The new method was compared with an existing commercialized image processing system, 'A' Product, for performance reliability tests. Also, the new method improved and developed the comparative advantage image processing algorithm. With regard to the test results, the algorithm over-counted one more vehicle, while 'A' product missed four vehicles during the test period. The first test results were used to improve the performance of this algorithm, and it's performance was improved though the second test in various and complicated traffic environment to yield superior performance.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.797-800
• /
• 2003

In the current study, the authors propose a method for extracting license plate regions by means of a neural network trained to output the plates center of gravity. The method is shown to be effective. Since the learning pattern presentation positions are defined by random numbers, a different pattern is submitted to the neural network for learning each time, which enables it to form a neural network with high universality of coverage. The article discusses issues of the optimal learning surface for a license plate revered by the learning pattern, the effort of suppression learning of the number and headlight sections, as well as the effect of learning pattern enlargement/reduction and of concentration value conversion. Results of evaluation tests based on pictures of 595 vehicles taken at an underground parking garage demonstrated detection rates of 98.5%.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.1 no.3
• /
• pp.105-109
• /
• 2008

In the current study, the authors propose a method for extracting license plate regions by means of a neural network trained to output the plate.s center of gravity. The method is shown to be effective. Since the learning pattern presentation positions are defined by random numbers, a different pattern is submitted to the neural network for learning each time, which enables it to form a neural network with high universality of coverage. The article discusses issues of the optimal learning surface for a license plate covered by the learning pattern, the effect of suppression learning of the number and headlight sections, as well as the effect of learning pattern enlargement/reduction and of concentration value conversion. Results of evaluation tests based on pictures of 595 vehicles taken at an underground parking garage demonstrated detection rates of 98.5%.

• Journal of the Korea Safety Management & Science
• /
• v.25 no.4
• /
• pp.67-72
• /
• 2023

Human and material damage can be reduced if the risk is evaluated by engineering analysis of fire combustion products, smoke concentration, and smoke movement speed in the event of a fire in apartment houses and officetels. In this study, a lot of research on related safety evaluation in the basement needs to be studied and reflected in design, so experimental research was conducted to analyze the flow of smoke through computer simulation and provide analysis data through evacuation safety evaluation. First of all, the five-story underground parking lot subject to simulation has a large floor area, which is advantageous for improving evacuation safety performance, but it uses temperature detectors to increase detection time and fire spread speed. Second, it was analyzed that the evacuation time at all evacuation ports did not exceed the evacuation time, and as the time from the start of evacuation to the evacuation time was 216.9% compared to the travel time, it was evaluated that the safety performance of the evacuation was secured. Third, the above simulation results are a comprehensive safety evaluation based on the non-operation of fire extinguishing facilities in the fire room to increase safety, which means that smoother evacuation safety performance can be secured by linking fire extinguishing facilities.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.45 no.3
• /
• pp.12-19
• /
• 2008

Detection of a number plate consists of three stages; division of a number plate, extraction of each character from the plate, recognition of the characters. Among of these three states, division stage of a number plate is the most important part and also the most time-consuming state. This paper suggests an effective region extraction method of a number plate for various images obtained from unmanned inspection systems of illegal parking violation, especially when we have to consider the diverse surrounding environments of roads. Our approaching method detects each region by investigating the characteristics in changes of brightness and intensity between the background part and character part, and the characteristics on character parts such as the sizes, heights, widths, and distance in between two characters. The method also divides a number plate into different types of the plate. This research can solve the number plate region detection failure problems caused by plate edge damages not only for Korean domestic number plates but also for new European style number plates. The method also reduces the time consumption by processing the detection in real-time, therefore, it can be used as a practical solution.

• Spatial Information Research
• /
• v.16 no.1
• /
• pp.19-32
• /
• 2008

Usually in South Korea, land cover type and topographic undulation are frequently changed even in a narrow area. However, most of researches using aerial LIDAR(LIght Detection And Ranging) data in abroad had been acquired in the study areas to be changed infrequently. This research was performed to explore reconstruction methodologies of 3D surface models considering the distribution of land cover type and topographic undulation. Composed of variously undulatory forests, rocky river beds and man-made land cover such as streets, trees, buildings, parking lots and so on, an area was selected for the research. First of all, the area was divided into three zones based on land cover type and topographic undulation using its aerial ortho-photo. Then, aerial LIDAR data was clipped by each zone and different 3D modeling processes were applied to each clipped data before integration of each models and reconstruction of overall model. These kinds of processes might be effectively applied to landscape management, forest inventory and digital map composition. Besides, they would be useful to resolve less- or over-extracted problems caused by simple rectangle zoning when an usual data processing of aerial LIDAR.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.10
• /
• pp.2461-2469
• /
• 1997

Target detection and recognition problems, in which neural networks are widely used, require translation invariant and real-time processing in addition to the requirements that general pattern recognition problems need. This paper presents a novel architecture that meets the requirements and explains effective methodology to train the network. The proposed neural network is an architectural extension of the shared-weight neural network that is composed of the feature extraction stage followed by the pattern recognition stage. Its feature extraction stage performs correlational operation on the input with a weight kernel, and the entire neural network can be considered a nonlinear correlation filter. Therefore, the output of the proposed neural network is correlational plane with peak values at the location of the target. The architecture of this neural network is suitable for implementing with parallel or distributed computers, and this fact allows the application to the problems which require realtime processing. Net training methodology to overcome the problem caused by unbalance of the number of targets and non-targets is also introduced. To verify the performance, the proposed network is applied to detection and recognition problem of a specific automobile driving around in a parking lot. The results show no false alarms and fast processing enough to track a target that moves as fast as about 190 km per hour.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.1
• /
• pp.153-159
• /
• 2012

Driver's view has blind spot of automobile surroundings due to physical components of automobile architecture. Obstacles on blind spot are the cause of car destruction and car accidents. Cars which produced in recent have obstacle detection sensors and rear view cameras which provide information of obstacles on the blind sopt, and have also AVM(Around View Monitoring) which provides automobile surroundings for driver's safe driving. During a low-speed travel while parking or moving in a narrow street, a driver get help for safe driving by taking information of automobile surroundings using the above-mentioned devices. In this paper, we present a design and implementation of a 4-sided monitoring (4SM) system, which helps a driver see an integrated view of a vehicle's perimeter at a glance, using a car PC connected to four cameras installed on the front, rear, left, and right sides.

• The KIPS Transactions:PartB
• /
• v.14B no.1 s.111
• /
• pp.13-22
• /
• 2007

In this paper, we propose an audience counting method in an auditorium based on pattern comparison. The previous counting methods based on object detection can't exactly count the audience in real time because auditorium has coarse illumination condition and so many audiences. Therefore, in this paper, we count the audience in an auditorium with fixed seats by the method which the pattern from each reference seat is compared to the pattern from each input seat. Especially, to overcome limitations based on either illumination or noise, two pattern comparison methods are efficiently employed and combined. One is based on the amplitude projection, and the other is based on Walsh-Hadamard Kernel. Walsh-Hadamard Kernel has the characteristic which complements amplitude projection. Therefore, we ran achieve the accurate counting in the presence of coarse illumination and noise. The experimental results show that our method performs well on sequences of images acquired in an auditorium. We also verify a realistic possibility for other applications applying our method to the parking positioning system.