• Journal of Advanced Navigation Technology
• /
• v.1 no.1
• /
• pp.35-46
• /
• 1997

Over last decade Multiple Target Tracking (MTT) has been the subject of numerous presentations and conferences [1979-1900]. Various approaches have been proposed to solve the problem. Representative works in the problem are Nearest Neighbor (NN) method based on non-probabilistic data association (DA), Multiple Hypothesis Test (MHT) and Joint Probabilistic Data Association (JPDA) as the probabilistic approaches. These techniques have their own advantages and limitations in computational requirements and in the tracking performances. In this paper, the three promising algorithms based on the NN standard filter, MHT and JPDA methods are presented and their performances against simulated multiple maneuvering targets are compared through numerical simulations.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.4 no.2
• /
• pp.184-191
• /
• 1999

일반적으로 시스템 인식과 제어를 위해 이용하는 다층망 신경회로망은 기존의 역전파알고리즘을 이용한다. 그러나 결선강도에 대한 오차의 기울기를 구하는 방법이기 때문에 국부적 최소점에 빠지기 쉽고, 수렴속도가 매우 늦으며 초기결선강도 값들이나 학습계수에 민감하게 반응한다. 이와 같은 단점을 개선하기 위해 본 논문에서는 칼만필터링 기법을 도입하여 수렴속도를 빠르게 하고 초기 결선강도의 영향을 받지 않도록 개선하였으며, 유도전동기의 속도추정과 제어에 적용하여 좋은 결과를 보였다.

• Journal of Digital Contents Society
• /
• v.13 no.3
• /
• pp.263-269
• /
• 2012

Grouping of characters in a 3D game can be very effective to play. An 3D game characters grouping method is proposed using Object Feature vector depending on their characteristics. In the case of NMT, the constitution of pattern is regular and directive. But MT is not. Such characteristic is extracted using Gabor Filter, then character is grouped. Through experiment, we obtain accuracy of more than 80% for grouping method using each feature. Thus, using this property, characters could be grouped effectively and it draws the game more speed and strategic actions as a result.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.835-838
• /
• 2003

Computer vision and natural-language dialogue play an important role in friendly human-machine interfaces for service robots. In this paper we describe an integrated face detection and face recognition system for a welfare robot, which has also been combined with the robot's speech interface. Our approach to face detection is to combine neural network (NN) and genetic algorithm (GA): ANN serves as a face filter while GA is used to search the image efficiently. When the face is detected, embedded Hidden Markov Model (EMM) is used to determine its identity. A real-time system has been created by combining the face detection and recognition techniques. When motivated by the speaker's voice commands, it takes an image from the camera, finds the face inside the image and recognizes it. Experiments on an indoor environment with complex backgrounds showed that a recognition rate of more than 88% can be achieved.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.11a
• /
• pp.133-134
• /
• 2018

MPEG-NNR (Compressed Representation of Neural Networks) aims to define a compressed and interoperable representation of trained neural networks. In this paper, a compressed representation of NN and its evaluation performance along with use cases of image/video compression in MPEG-NNR are presented. In the compression of NN, a CNN to replace the in-loop filter in VVC (Versatile Video Coding) intra coding is compressed by applying uniform quantization to reduce the trained weights, and the compressed CNN is evaluated in terms of compression ratio and coding efficiency compared to the original CNN. Evaluation results show that CNN could be compressed to about quarter with negligible coding loss by applying simple quantization to the trained weights.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.1
• /
• pp.75-80
• /
• 2010

Noise handing is very important in voice recognition of vehicle environment. that has been studying about to hardware and software approach. hardware method that is noise filter circuit design, basically using Low-pass filter. it was shown a good result. and the side of software that has been developing about to algorithm for Noise canceler, NN(neural network), etc. in this paper we have analysis about to classified parameter pattern matting level for voice recognition on car noise environment that use of DTW(Dynamic Time Warping) which is applicable time series pattern recognition algorithm.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.15 no.2
• /
• pp.102-110
• /
• 2015

The main source of noise in computed tomography (CT) images is a quantum noise, which results from statistical fluctuations of X-ray quanta reaching the detector. This paper proposes a neural network (NN) based hybrid filter for removing quantum noise. The proposed filter consists of bilateral filters (BFs), a single or multiple neural edge enhancer(s) (NEE), and a neural filter (NF) to combine them. The BFs take into account the difference in value from the neighbors, to preserve edges while smoothing. The NEE is used to clearly enhance the desired edges from noisy images. The NF acts like a fusion operator, and attempts to construct an enhanced output image. Several measurements are used to evaluate the image quality, like the root mean square error (RMSE), the improvement in signal to noise ratio (ISNR), the standard deviation ratio (MSR), and the contrast to noise ratio (CNR). Also, the modulation transfer function (MTF) is used as a means of determining how well the edge structure is preserved. In terms of all those measurements and means, the proposed filter shows better performance than the guided filter, and the nonlocal means (NLM) filter. In addition, there is no severe restriction to select the number of inputs for the fusion operator differently from the neuro-fuzzy system. Therefore, without concerning too much about the filter selection for fusion, one could apply the proposed hybrid filter to various images with different modalities, once the corresponding noise characteristics are explored.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.321-324
• /
• 2002

In this paper, we introduce a new approach to fingerprint classification based on both singular points and gabor features. We find singular points of fingerprint image by using squared direction field and Poincare index. Then, the input fingerprint image can be classified into one of 5 classes using the number of singular points and their location. However, it is often impossible to classify the fingerprint image because the numbers and the position of the singular points are not correct due to noise. In this case Gabor features are extracted from unclassified images using Gator filter and they are classified by using k-NN classifier. This method has been tested on the NIST-4 database. The experimental results show that the proposed method is reliable.

• Journal of Positioning, Navigation, and Timing
• /
• v.9 no.3
• /
• pp.285-292
• /
• 2020

This paper introduces a new computational efficient Dilution of Precision (DOP)-based landmark exclusion method while ensuring the safety of the LiDAR-based navigation system that uses an innovation-based Nearest-Neighbor (NN) Data Association (DA) process. The NN DA process finds a correct landmark association hypothesis among all potential landmark permutations using Kalman filter innovation vectors. This makes the computational load increases exponentially as the number of landmarks increases. In this paper, we thus exclude landmarks by introducing DOP that quantifies the geometric distribution of landmarks as a way to minimize the loss of integrity performance that can occur by reducing landmarks. The number of landmarks to be excluded is set as the maximum number that can satisfy the integrity risk requirement. For the verification of the method, we developed a simulator that can analyze integrity risk according to the landmark number and its geometric distribution. Based on the simulation, we analyzed the relationship between DOP and integrity risk of the DA process by excluding each landmark. The results showed a tendency to minimize the loss of integrity performance when excluding landmarks with poor DOP. The developed method opens the possibility of assuring the safety risk of the Lidar-based navigation system in real-time applications by reducing a substantial amount of computational load.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.55 no.9
• /
• pp.401-408
• /
• 2006

In a multi-sensor multi-target tracking systems, the local sensors have the role of tracking the target and transferring the measurements to the fusion center. The measurements from the same target can arrive out of sequence called the out-of-sequence measurements(OOSMs). Out-of-sequence measurements can arise at the fusion center due to communication delay and varying preprocessing time for different sensor platforms. In general, the track fusion occurs to enhance the tracking performance of the sensors using the measurements from the sensors at the fusion center. The target informations can wive at the fusion center with the clutter informations in cluttered environment. In this paper, the OOSM update step with MPDA(Most Probable Data Association) is introduced and tested in several cases with the various clutter density through the Monte Carlo simulation. The performance of the MPDA with OOSM update step is compared with the existing NN, PDA, and PDA-AI for the air target tracking in cluttered and out-of-sequence measurement environment. Simulation results show that MPDA with the OOSM has compatible root mean square errors with out-of-sequence PDA-AI filter and the MPDA is sufficient to be used in out-of-sequence environment.