Papers by Keyword: Independent Component Analysis (ICA)

Paper TitlePage

Authors: Zuo Wei Huang, Shu Guang Wu, Tao Xin Zhang
Abstract: Hyperspectral remote sensing is the multi-dimensional information obtaining technology,which combines target detection and spectral imaging technology together, In order to accord with the condition of hyperspectral imagery,the paper developed an optimized ICA algorithm for change detection to describe the statistical distribution of the data. By processing these abundance maps, change of different classes of objects can be obtained..A approach is capable of self-adaptation, and can be applied to hyperspectral images with different characteristics. Experiment results demonstrate that the ICA-based hyperspectral change detection performs better than other traditional methods with a high detection rate and a low false detection rate.
Authors: Gao Ling, Shou Xin Ren
Abstract: A multi-dimensional data processing method, independent component analysis-based principal component regression (ICA-PCR) was developed for simultaneous kinetic determination of Cu (II), Fe (III) and Ni (II). Independent component analysis is a newly developed signal processing technique aiming at solving related blind source separation (BSS) problem. One program, PICAPCR, was designed to perform relative calculations. Experimental results showed the ICA-PCR method to be successful for simultaneous multicomponent kinetic determination even where there was severe overlap of spectra.
Authors: Xin Zhang, Ting Fang
Abstract: Quantitative assessment of muscle spasticity is an important issue in the rehabilitation medicine. However, there are a small amount of qualitative and quantitative assessment be used in clinical studies for that the mechanism of vasospasm is not clear yet. The objective of this paper is to suggest a totally new method using independent component analysis (ICA) to analyze spasticity and indicate possibility to apply this method for evaluation of spasticity. In order to simulate the experiment of human swing leg angle bending system, an entry system with the input signal and the original angle had been established. And the system's pulse signal had been reconstructed. By comparing the thigh swing test experiments of healthy people and patients and analyzing their different impulse signals and impulse responses, the index of evaluation of spasticity had been obtained.
Authors: Hai Dong Guo, Shun Ming Li, Yuan Yuan Zhang, Xing Xing Wang, Sai Ma
Abstract: For weak vibration signal with strong noise, a new kind of weak vibration signal detection method was proposed in this paper. Based on the redundancy reducing capability and the uncertain amplitude of independent component analysis, virtual noise was introduced to extend the dimension of original observed signal after we analyzed the prior features of noises in observed signal. Then extended signals were processed to get the independent source signals by applying to blind source separation (BSS). Thus, the noise embedded in observed signal was removed and characteristics of weak vibration signal were obtained successfully. Through the theoretical analysis and the simulation, the introduced method of this paper was checked to be available and then it was applied to faults analysis of rotor misalignment successfully. Finally, we made a conclusion that this method had great application value for the extraction of weak vibration signal.
Authors: Shuang Wei, De Fu Jiang, Yang Gao
Abstract: This paper presents a diversity-guided Particle swarm optimization (PSO) algorithm to resolve the Blind source separation (BSS) problem. Because the independent component analysis (ICA) approach, a popular method for the BSS problem, has a shortcoming of premature convergence during the optimization process, the proposed PSO algorithm aims to improve this issue by using the diversity calculation to avoid trapping in the local optima. In the experiment, the performance of the proposed PSO algorithm for the BSS problem has been investigated and the results are compared with the conventional PSO algorithm. It shows that the proposed PSO algorithm outperforms the conventional PSO algorithm.
Authors: Wei Yu, Qiang Han, Jing Jing Ma, Pei Xie
Abstract: Faint signal extraction is always a difficult issue in biomedical signal processing field, because the desired signal is often submerged in several relatively large signals or noises. A novel faint signal processing method based on Empirical Mode Decomposition (EMD) and Independent Component Analysis (ICA) is developed to enhance the sensitivity and reliability of faint signal detection. This novel method includes two major steps, which is, firstly the decomposition of the biomedical composite signal using EMD, then the classification or extraction of the desired faint signal component through ICA. This paper explored the working principles and the performance of this novel signal processing method under the specific biomedical environment of fetal electrocardiogram extraction (FECG). The experimental results show that the proposed method has better extraction effect and quality compared with traditional ICA methods.
Authors: Yu Feng Xue, Yu Jia Wang, Qiu Dong Sun
Abstract: In this paper, a new method is introduced to derive the extended natural gradient, which was proposed by Lewicki and Sejnowski in [1]. However, they made their derivation under many approximations, and the proof is also very complicated. To give a more rigors mathematical proof for this gradient, the Lie group invariance property is introduced which makes the proof much easier and straightforward. In addition, an iterative algorithm through Newton's method is also given to estimate the sources efficiently. The results of the experiments confirm the efficiency of the proposed method.
Authors: Lei Feng, Xiao Fei Shi, Hong Yu Chen, Yan Hua Li, Yue Long Zhang
Abstract: Most existing watermark extraction algorithms were dependent on prior knowledge. This paper proposed a blind extraction method without relying on prior knowledge. According to constructing new observation based on nonsubsampled contourlet transform, which utilizes low frequency and directional components of watermarked image, more independent components are generated. We involve these components into watermarked image and resort this solution to multichannel blind source separation. Estimated watermark is recovered by ICA algorithm. Experiment results indicate that the proposed method can achieve better results in contrast with two existing algorithms.
Authors: Hao Cheng Wu, Yong Shou Dai, Wei Feng Sun, Li Gang Li, Ya Nan Zhang
Abstract: Periodic noise is an important manifestation of the drill string vibration signal noise. In order to extract the characteristics of the signals which reflect the situation of the tools in drilling, the periodic components which influence the original drill string vibration signal in the well field were researched and the independent component analysis algorithm which is on the basis of negative entropy for periodic vibration noise separation was adopted. At the same time, the effect of algorithm demixing was improved where periodic noise components which existed in three directions of drill string vibration signals were used, combining with the improved particle swarm optimization algorithm to seek the optimal mixed matrix by which the multi-channel mixed-signal of independent component analysis algorithm could be structured. This method in operation was fast. And after separation each signal was of high similarity. Through the experimental simulation, the method was proven effective in the drill string vibration periodic noise signal separation.
Authors: Arjon Turnip, Grace Gita Redhyka, Hilman S. Alam, Iwan R. Setiawan
Abstract: In this paper, an experiment of spike detection based mental task with ayes movement stimuli is reported. The approximation of ICA algorithm is required to eliminate artifacts and detect a pike of brain activity according to the given stimuli which are normal, closed, and blinking ayes. A comparison of ICA algorithms based Extended Fourth Order Blind Identification and Algorithm for Multiple Unknown Signal Extraction is tested. The quality of the extracted signals is measured through the value of the signal to interference ratio and signal to distortion ratio. The extracted results indicate that the best spike detection is achieved using AMUSE algorithm.Keywords : EEG , s pike , Independent Component Analysis (ICA).
Showing 1 to 10 of 76 Paper Titles