• Journal of Power System Engineering
• /
• v.5 no.2
• /
• pp.93-99
• /
• 2001

This paper presents the automatic recognition system for car number plate. In our country, two types of number plate pattern is used. The one is old type of number plate, the other is new type of number plate. To recognize both new and old type number plates, the system must have flexibility. Therefore, in this paper, automatic recognition system is developed by use of the neural network for good adaptation, good generalization, and modulation. And because the number plate is made of three codes, the multi neural network consists of three networks. Neural network is teamed by GDR(Generalized Delta learning Rule) and it is verified the effectiveness of the method through experimental results.

• Phonetics and Speech Sciences
• /
• v.2 no.4
• /
• pp.145-151
• /
• 2010

In this paper, a robust feature compensation method is proposed for improving the performance of speech recognition. The proposed method is incorporated into the auto-regressive moving average (ARMA) based feature compensation. We employ variable weights for the ARMA filter according to the degree of speech activity, and pass the normalized cepstral sequence through the weighted ARMA filter. Additionally when normalizing the cepstral sequences in training, the cepstral means and variances are estimated from total training utterances. Experimental results show the proposed method significantly improves the speech recognition performance in the noisy and reverberant environments.

• Smart Media Journal
• /
• v.8 no.3
• /
• pp.47-52
• /
• 2019

In this paper, we designed HW / SW interfaces for processing the signals of capacitive sensors like Electric Potential Sensor (EPS) to detect the surrounding electric field disturbance as feature signals in motion recognition systems. We implemented a smart light control system with those interfaces. In the system, the on/off switch and brightness adjustment are controlled by hand gestures using the designed and fabricated interface circuits. PWM (Pulse Width Modulation) signals of the controller with a driver IC are used to drive the LED and to control the brightness and on/off operation. Using the hand-gesture signals obtained through EPS sensors and the interface HW/SW, we can not only construct a gesture instructing system but also accomplish the faster recognition speed by developing dedicated interface hardware including control circuitry. Finally, using the proposed hand-gesture recognition and signal processing methods, the light control module was also designed and implemented. The experimental result shows that the smart light control system can control the LED module properly by accurate motion detection and gesture classification.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.3
• /
• pp.427-432
• /
• 2021

In this paper, we conduct performance analysis in automatic modulation classification of unknown communication signal to identify its modulation types based on deep neural network. The modulation classification performance was verified using time domain digital sample data of the modulated signal, frequency domain data to which FFT was applied, and time and frequency domain mixed data as neural network input data. For 11 types of analog and digitally modulated signals, the modulation classification performance was verified in various SNR environments ranging from -20 to 18 dB and reason for false classification was analyzed. In addition, by checking the learning speed according to the type of input data for neural network, proposed method is effective for constructing an practical automatic modulation recognition system that require a lot of time to learn.

• Journal of Broadcast Engineering
• /
• v.22 no.2
• /
• pp.257-260
• /
• 2017

In this letter, we propose an efficient automatic modulation recognition (AMR) method which classifies digitally modulated signals by estimating the bandwidth. In AMR, feature-based methods are widely used and the accuracy of the features is highly dependent on the number of symbols and the number of samples per symbol (NSPS). In this letter, at first, we coarsely estimate the bandwidth of the oversampled signals, and then decrease the sample rate to yield adequate NSPS. As a result, more symbols are used for AMR and the correct classification rate becomes high under the same number of samples.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.24 no.12
• /
• pp.1173-1180
• /
• 2013

High Resolution Range Profile-Jet Engine Modulation imagery first introduced in 2005 carries out radar target recognition by localizing the position of the jet engine installed on the aircraft target. This paper presents a new approach for estimating the jet engine position in the HRRP-JEM image based on the eccentricity of a complex signal. It can effectively evaluate the contribution of the JEM component to the radar received signal in a range bin of the HRRP-JEM image. Therefore, the localization is expected to be performed more quantitatively and reliably by pinpointing the range bin corresponding to the jet engine position where the JEM contribution is maximized. The simulation results of realistic aircraft models validated the effectiveness of the proposed concept.

• The Journal of the Korea Contents Association
• /
• v.4 no.4
• /
• pp.205-213
• /
• 2004

In this paper, we proposed a wavelet phase-only filter modulation synthetic discriminant function joint transform correlator(WPOfSDF-JTC) for scale invariant pattern recognition, and an improved algorithm to reduce the filter synthesis time. Computer simulation showed that the proposed filter has better SNR than CWMF if input image has random noise and the improved synthesis algorithm can reduce the iteration time. We used frequency selective JTC to solve the problem of the optical alignment and eliminate the autocorrelation and crosscorrelation between each input image.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.12 no.1
• /
• pp.35-41
• /
• 2017

We propose a face recognition and notification system that can transform visual face information into tactile signals in order to help visually impaired people. The proposed system consists of a glasses type camera, a mobile computer and an electronic cane. The glasses type camera captures the frontal view of the user, and sends this image to mobile computer. The mobile computer starts to search for human's face in the image when obstacles are detected by ultrasonic sensors. In a case that human's face is detected, the mobile computer identifies detected face. At this time, Adaboost and compressive sensing are used as a detector and a classifier, respectively. After the identification procedures of the detected face, the identified face information is sent to controller attached to a cane using a Bluetooth communication. The controller generates motor control signals using Pulse Width Modulation (PWM) according to the recognized face labels. The vibration motor generates vibration patterns to inform the visually impaired person of the face recognition result. The experimental results of face recognition and notification system show that proposed system is helpful for visually impaired people by providing person identification results in front of him/her.

• Journal of Advanced Navigation Technology
• /
• v.28 no.4
• /
• pp.428-435
• /
• 2024

In this paper, a human activity regeneration (HAR) system based on multiple input multiple output frequency modulation continuous wave (MIMO FMCW) radar was designed and implemented. Using point cloud data from MIMO radar sensors has advantages in terms of privacy, safety, and accuracy. For the implementation of the HAR system, a customized neural network based on PointPillars and depthwise separate convolutional neural network (DS-CNN) was developed. By processing high-resolution point cloud data through a lightweight network, high accuracy and efficiency were achieved. As a result, the accuracy of 98.27% and the computational complexity of 11.27M multiply-accumulates (Macs) were achieved. In addition, the developed neural network model was implemented on Raspberry-Pi embedded system and it was confirmed that point cloud data can be processed at a speed of up to 8 fps.

• Korean Journal of Optics and Photonics
• /
• v.20 no.2
• /
• pp.76-80
• /
• 2009

Finger-print recognition is achieved by comparing an input finger-print image with the stored images in the computer, and finally by determining agreement or disagreement. Encryption and decryption are necessary in the finger-print recognition process. In these process CGH (Computer Generated Hologram) is used, and finger-print images reconstructed from the CGHs are compared. In this paper, two methods of recognition are used, one is to compare the finger-print images of each other reconstructed from their CGHs and the other is to compare the CGHs to each other directly, to analyze the differences of finger-print recognition capability between these two methods. Experimental results show that the capability of finger-print recognition for comparing the CGHs of the two is about 150 times higher than in case of comparing the reconstructed finger-print images. Especially the changes of characteristics according to modulation types of CGH are analyzed.