• Journal of Information Processing Systems
• /
• v.19 no.6
• /
• pp.778-790
• /
• 2023

Although density peak clustering can often easily yield excellent results, there is still room for improvement when dealing with complex, high-dimensional datasets. One of the main limitations of this algorithm is its reliance on geometric distance as the sole similarity measurement. To address this limitation, we draw inspiration from the information bottleneck theory, and propose a novel density peak clustering algorithm that incorporates this theory as a similarity measure. Specifically, our algorithm utilizes the joint probability distribution between data objects and feature information, and employs the loss of mutual information as the measurement standard. This approach not only eliminates the potential for subjective error in selecting similarity method, but also enhances performance on datasets with multiple centers and high dimensionality. To evaluate the effectiveness of our algorithm, we conducted experiments using ten carefully selected datasets and compared the results with three other algorithms. The experimental results demonstrate that our information bottleneck-based density peaks clustering (IBDPC) algorithm consistently achieves high levels of accuracy, highlighting its potential as a valuable tool for data clustering tasks.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.10
• /
• pp.1826-1835
• /
• 2007

In this paper, we present a new approach for automatic detection and tracking for multiple targets. We combine a highest probability data association(HPDA) algorithm for target detection with a particle filter for multiple target tracking. The proposed approach evaluates the probabilities of one-to-one assignments of measurement-to-track and the measurement with the highest probability is selected to be target- originated, and the measurement is used for probabilistic weight update of particle filtering. The performance of the proposed algorithm for target tracking in clutter is compared with the existing clustering algorithm and the sequential monte carlo method for probability hypothesis density(SMC PHD) algorithm for multi-target detection and tracking. Computer simulation studies demonstrate that the HPDA algorithm is robust in performing automatic detection and tracking for multiple targets even though the environment is hostile in terms of high clutter density and low target detection probability.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.1
• /
• pp.37-44
• /
• 2009

Density is the most important congestion indicator among the three fundamental flow variables, flow, speed and density. Measuring density in the field has two different ways, direct and indirect. Taking photos with wide views is one of direct ways, which is not widely used because of its cost and lacking of proper positions. Another direct density measuring method using two spot detectors has been introduced with the concept of instantaneous density, average density and measurement interval. The relationship between accuracy and measurement interval has been investigated using the simulation data produced by Paramics API function. Finally, density measurement algorithm has been suggested including exponential smoothing for device development.

• Journal of the Korean Society for Precision Engineering
• /
• v.21 no.11
• /
• pp.200-208
• /
• 2004

In ultrasonic bone densitometry, the positioning of measurement site is decisive in precision and reproducibility. In this study, automatic Region of Interest (ROI) detection algorithm is suggested and adopted the method using the local minimum value by ultrasonic image. The preprocess before the local minimum method extracts out the bone area and calculates the geometrical information of bone. The developed ROI detection algorithm was applied to the clinical test for the subject of 305 female patients in the range of 22-88 years old. As the results, the accuracy of the algorithm was shown to be 98.3％. It was also found that bone density parameter was significantly correlated with age(r=0.85, p<0.0001).

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.4
• /
• pp.354-360
• /
• 2013

In this paper, it is discussed that the detection and tracking performance of the piecewise linear path underwater target is improved using clutter reduction algorithm in heavy clutter density environment. Through clutter reduction algorithm using Hough Transform, measurements which represent clutter features are removed and the performance of target tracking on the remaining measurements is demonstrated applying CMKF-L(Converted Measurement Kalman Filter with Linearization) as tracking filter. Algorithm performance test is conducted using simulation data and real sea-trial data and by applying the proposed algorithm in heavy clutter density environment, it is confirmed that the target is tracked consistently and stably with clutter rejected measurements.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 1996.06c
• /
• pp.792-801
• /
• 1996

A thinning algorithm -based image analysis technique was developed to measure corn root lengths. The root length measurement method was evaluated by comparing thread lengths measured by the image analysis system with actual thread lengths. The length measurement method accurately estimated actual thread lengths (less than 2% calculated error). Also, a rapid root length density measurement procedure, which utilizes the above root length measurement method, was developed to estimate corn root length density without washing the roots. Root length densities estimated from the cut soil surface of core samples taken from the field were paired with the root length densities determined from washed roots from the same soil core sample. A linear relationship between these two values was expected and was found. Eliminating the root washing procedure reduces the time required for measuring corn root length density substantially.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.10
• /
• pp.5095-5111
• /
• 2016

The Probability Hypothesis Density (PHD) filter is a suboptimal approximation and tractable alternative to the multi-target Bayesian filter based on random finite sets. However, the PHD filter fails to track newborn targets when the target birth intensity is unknown prior to tracking. In this paper, a dual detection-guided newborn target intensity PHD algorithm is developed to solve the problem, where two schemes, namely, a newborn target intensity estimation scheme and improved measurement-driven scheme, are proposed. First, the newborn target intensity estimation scheme, consisting of the Dirichlet distribution with the negative exponent parameter and target velocity feature, is used to recursively estimate the target birth intensity. Then, an improved measurement-driven scheme is introduced to reduce the errors of the estimated number of targets and computational load. Simulation results demonstrate that the proposed algorithm can achieve good performance in terms of target states, target number and computational load when the newborn target intensity is not predefined in multi-target tracking systems.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.29 no.10 s.241
• /
• pp.1430-1437
• /
• 2005

It is very important to achieve a high reproducibility in the ultrasonic measurement of bone mineral density. In this study, we examined number of sampling waveform, control of temperature, diameter of region of interest as factors to improve reproducibility. We decided the optimal number of waveforms to be converted to frequency domain as period of 1. We have minimized the effects of variable temperature and constrained generation of micro bubble by keeping temperature within a range of $32\pm0.5^{\circ}C$ with a precise temperature controlling algorithm. We also found the optimal diameter of region of interest to be 13mm. In this paper, we demonstrated the improved reproducibility by controlling various factors affecting the ultrasonic measurement of bone mineral density.

• Journal of Korean Society of Transportation
• /
• v.28 no.3
• /
• pp.157-167
• /
• 2010

Density is the most important congestion indicator among the three fundamental flow variables, flow, speed and density. Measuring density in the field has two different ways, direct and indirect. Taking photos with wide views is one of direct ways, which is not widely used because of its cost and lacking of proper positions. Another direct density measuring method using two point detectors has been introduced with the concept of instantaneous density, average density and measurement interval. The relationship between accuracy and measurement interval has been investigated using the SIMULATION data produced by Paramics Application Programming Interface function. We analyze the affect of segment density accuracy by sensing gap each road condition such as sensing segment length, lane and LOS after gathering data by Paramics Application Programming Interface.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.7
• /
• pp.40-48
• /
• 2019

New concept of fine dust measurement method is suggested based on ultrasonic scattering. These days, fine dust has been social problem in Korea, and many researches has been conducted including the area structural maintenance. Conventional measurement system such as optical scattering and semiconductor has a limit from environmental factors like relative humidity. However, ultrasound is based on mechanical waves, which perturb mechanical properties of medium such as density and elastic constants. Using the advantage, the algorithm for fine dust measurement is derived and evaluated using 2-D finite difference method. The numerical analysis simulates ultrasonic wave propagation inside multiple scattering medium like fine dust in air. Signal processing scheme is also suggested and the results show that the error of the algorithm is around minimum of 0.7 and maximum of 24.9 in the number density unit. It is shown that cross-section of fine dust is a key parameter to improve the accuracy of algorithm.