Article
Computer Science, Software Engineering
Guang Yi Chen, Adam Krzyzak, Shen-En Qian
Summary: In this paper, a new method for HSI classification is proposed, which combines MNF, SF, and SVM to consider both spatial and spectral information. Experimental results demonstrate the competitiveness of the proposed method against existing classification methods.
INTERNATIONAL JOURNAL OF WAVELETS MULTIRESOLUTION AND INFORMATION PROCESSING
(2022)
Article
Instruments & Instrumentation
Xin Huang, Qinghua Meng, Zhefeng Wu, Fuxian He, Pan Tian, Jiaojiao Lin, Hui Zhu, Xulin Zhou, Yuqing Huang
Summary: This paper explores the early detection and extraction of bruised regions on Gongcheng persimmons using hyperspectral imaging techniques. Hyperspectral images of Gongcheng persimmons in the visible and near-infrared regions were acquired, and the effects of different preprocessing methods on the accuracy of the support vector machine classification model were compared. A genetic algorithm was used to optimize the model parameters, and a successive projections algorithm was used to select the best wavelength for identifying the bruised parts.
INFRARED PHYSICS & TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Qi Xie, Jianping Hu, Xiaochao Wang, Ying Du, Hong Qin
Summary: This paper presents several optimization-based approaches to improve bidimensional empirical mode decomposition (BEMD) for multi-scale feature description of input images. The proposed approaches, including improved unconstrained optimization (IUOA-BEMD), scale-guided optimization (SGO-BEMD), and edge-aware scale-guided optimization (EASGO-BEMD), have shown competitive performance in visualizing and quantitatively analyzing various images. Furthermore, the integration of BEMD with Retinex theory for image contrast and brightness enhancement represents a novel attempt.
DIGITAL SIGNAL PROCESSING
(2023)
Proceedings Paper
Automation & Control Systems
Linlin Chen, Linzhao Hao, Fulong Liu, Quan Chen
Summary: Supervised classification is widely used in hyperspectral data analysis, but it faces great challenges due to the large number of data bands and redundancy between them. In this paper, a joint feature extraction method based on MNF and VMD is proposed to improve the classification accuracy. SVM classification and comparative analysis demonstrate the effectiveness of the proposed method.
2022 IEEE 6TH ADVANCED INFORMATION TECHNOLOGY, ELECTRONIC AND AUTOMATION CONTROL CONFERENCE (IAEAC)
(2022)
Article
Engineering, Multidisciplinary
S. Z. T. Motlagh, A. Akbari Foroud
Summary: This study proposes an approach to identify multiple flicker sources using voltage signals and various signal processing and feature selection techniques. The effectiveness of the proposed method is validated through simulation results.
Article
Materials Science, Multidisciplinary
Guang yi Chen, A. D. A. M. Krzyzak, Shen-en Qian
Summary: In this paper, we improve the PCA-based EPFs for HSI classification by using MNF instead of PCA to reduce the dimensionality. Our new method outperforms the PCA-based EPFs in noisy environments and even without noise. The experimental results validate the superiority of our MNF+EPFs approach in remote sensing.
IMAGE ANALYSIS & STEREOLOGY
(2023)
Article
Optics
Wen Huo, Chenxing Wang, Feipeng Da
Summary: This paper studies the BSEMD and iBSEMD algorithms, proposes a key factor to simplify fringe enhancement, and validates the efficiency and effectiveness of the method through experiments.
Article
Computer Science, Information Systems
Guang Yi Chen, Adam Krzyzak, Wen Fang Xie
Summary: This paper introduces a novel method for hyperspectral face recognition, incorporating HOG features, CRC classifier, and denoising techniques, showing improved performance through experiments.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Chemistry, Analytical
Chun-Hsien Hsu, Ya-Ning Wu
Summary: This paper introduces the application of empirical mode decomposition on MEG data, showing that improved data clarity can capture distinguishing features between conditions and raise interesting questions about hemispheric dominance in the encoding of facial and identity information.
Article
Automation & Control Systems
Wenjian Li, Nian Cai, Zhou Ning, Yongchao Dong, Han Wang
Summary: A nonlinear error compensation method based on empirical mode decomposition and adaptive intrinsic mode function selection is proposed to improve the measurement precision of the optical encoder.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2022)
Article
Materials Science, Characterization & Testing
Leonardo Oldani Felix, Dionisio Henrique Carvalho de Sa So Martins, Ulisses Admar Barbosa Vicente Monteiro, Brenno Moura Castro, Luiz Antonio Vaz Pinto, Carlos Alfredo Orfao Martins
Summary: Gearboxes are widely used in various industries, but identifying faults is challenging due to their complexity. This study used EMD and PCC to process vibration signals, and employed feature selection and machine learning algorithms for fault classification. The results showed that the ANN model performed the best.
JOURNAL OF NONDESTRUCTIVE EVALUATION
(2023)
Article
Plant Sciences
Zheli Wang, Wenqian Huang, Xi Tian, Yuan Long, Lianjie Li, Shuxiang Fan
Summary: In this study, hyperspectral reflectance imaging was used to identify aged maize seeds, and classification models were developed using support vector machine algorithm. The results showed that combining two wavelengths with image texture features could effectively detect aged seeds.
FRONTIERS IN PLANT SCIENCE
(2022)
Article
Physics, Multidisciplinary
Shuoxuan Dong, Yang Zhou, Tianyi Chen, Shen Li, Qiantong Gao, Bin Ran
Summary: This paper introduces a novel trajectory reconstruction method using Empirical Mode Decomposition (EMD) and Butterworth low-pass filter framework to reduce noise interference while maintaining physical integrity. Experimental results show that the method effectively removes high-frequency noise and maintains data integrity, producing speed and acceleration information of higher quality compared to existing methods.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2021)
Correction
Computer Science, Software Engineering
Xiaochao Wang, Kun Hu, Jianping Hu, Ling Du, Anthony T. S. Ho, Hong Qin
Summary: The article contained mistakes regarding the affiliation and biography of Ling Du, which have been corrected.
Article
Chemistry, Analytical
Selma Grebovic, Nermin Oprasic, Vahid Helac, Ivo Uglesic, Abdulah Aksamovic, Samim Konjicija
Summary: This paper introduces a direct lightning current measurement system installed on Mount Lovcen, and analyzes the observed data using Matlab. The method allows for obtaining statistical data on lightning parameters and proposes a classification and analysis approach based on empirical mode decomposition.
Article
Engineering, Geological
Zheng-Yi Feng, Shih-Hao Chen
Summary: This study discusses the seismic signals produced by soil ruptures during landslides and their potential for predicting landslides and understanding different landslide processes. Different types of seismic signals were recorded during sliding processes and three slide types were identified. Precursor signals before single slide events have potential warning applications.
Article
Environmental Sciences
Ming-Der Yang, Hsin-Hung Tseng, Yu-Chun Hsu, Chin-Ying Yang, Ming-Hsin Lai, Dong-Hong Wu
Summary: This paper presents a validated dataset of annotated UAV images, detailing data acquisition, preprocessing, and showcasing a CNN classification. The dataset includes a multi-rotor UAV platform flying a planned scouting route over rice paddies.
Article
Environmental Sciences
Kai-Yun Li, Niall G. Burnside, Raul Sampaio de Lima, Miguel Villoslada Pecina, Karli Sepp, Ming-Der Yang, Janar Raet, Ants Vain, Are Selge, Kalev Sepp
Summary: This study used machine learning techniques and multispectral vegetation indices to predict the dry matter yields of red clover-grass mixtures under different farming operations. The results showed the best performance of the artificial neural network model, which was influenced by farming operations.
Article
Engineering, Geological
Zheng-Yi Feng, Rui-Chia Zhuang
Summary: Seismic and acoustic signals induced by rock falls contain valuable information about movement processes. By monitoring and analyzing these signals simultaneously, behaviors of the processes can be closely interpreted. The study found that seismic and acoustic signals are generally similar and correlated, with acoustic signals attenuating more slowly than seismic signals. Additionally, the study estimated the surface wave velocity of the stratum and observed that larger rock masses had lower seismic frequencies.
Article
Chemistry, Analytical
Ming-Der Yang, Yu-Chun Hsu, Wei-Cheng Tseng, Chian-Yu Lu, Chin-Ying Yang, Ming-Hsin Lai, Dong-Hong Wu
Summary: The study utilizes smartphone images of rice panicles and machine learning models to achieve real-time and cost-effective measurement of grain moisture content, enabling on-farm prediction of harvest dates and scheduling of agricultural machinery. This method partially replaces traditional time-consuming testing methods.
Article
Environmental Sciences
Hone-Jay Chu, Regita Faridatunisa Wijayanti, Lalu Muhamad Jaelani, Hui-Ping Tsai
Summary: This study improved drought monitoring in Java, Indonesia using satellite precipitation data by establishing a standardized precipitation index and conducting spatial downscaling for higher accuracy. Spatial downscaling was found to be more suitable for heterogeneous SPI, especially during transitional periods, leading to more accurate results.
Article
Environmental Sciences
Hui Ping Tsai, Wei-Ying Wong
Summary: The study utilized 30 years of AVHRR NDVI3g monthly data to identify natural clusters and driving factors in the upstream watersheds of Taiwan, resulting in the identification of six clusters and the explanation of approximately 52% of NDVI variance by environmental factors.
Article
Environmental Sciences
Hui Ping Tsai, Geng-Gui Wang, Zhong-Han Zhuang
Summary: This study explored the long-term trends and breakpoints of vegetation, rainfall, and temperature in Taiwan from 1982 to 2012, revealing different patterns and breakpoints, especially in vertical differences. Regional variations showed stable vegetation growth in the north and worsening in the central region, with larger variations at higher elevations. There was a significant negative correlation between climate factors and NDVI, while temperature had positive effects at low altitudes below 500m.
Article
Environmental Sciences
Kai-Yun Li, Raul Sampaio de Lima, Niall G. Burnside, Ele Vahtmaee, Tiit Kutser, Karli Sepp, Victor Henrique Cabral Pinheiro, Ming-Der Yang, Ants Vain, Kalev Sepp
Summary: This study integrates autonomous computation and AI technologies with a hyperspectral system to estimate crop yield and biomass. The research shows the significant estimation capacity of the AutoML regression model and highlights the economic and environmental benefits of the hyperspectral system in sustainable and intelligent agriculture.
Article
Chemistry, Multidisciplinary
Chung-Jui Lai, Hui-Ping Tsai, Ju-Yu Chen, Mei-Xuan Wu, You-Jie Chen, Kun-Yi Lin, Hong-Ta Yang
Summary: This study presents a scalable and simple non-lithography-based approach to engineer robust antireflective structures inspired by bio-design. The biomimetic coating shows improved antireflection performance and has considerable technological importance in practical applications.
Article
Environmental Sciences
Hsin-Hung Tseng, Ming-Der Yang, R. Saminathan, Yu-Chun Hsu, Chin-Ying Yang, Dong-Hong Wu
Summary: This study focuses on detecting rice seedlings in paddy fields using transfer learning from UAV-acquired images. The results show that CNN-based models perform better than the traditional HOG-SVM approach. The adoption of transfer learning allows for the rapid establishment of object detection applications with promising performance.
Article
Plant Sciences
Dong-Hong Wu, Chung-Tse Chen, Ming-Der Yang, Yi-Chien Wu, Chia-Yu Lin, Ming-Hsin Lai, Chin-Ying Yang
Summary: This study investigated the relationship between rice lodging and various cultivation conditions. The results showed that lodging was closely related to nitrogen fertilizer content and plant height in the booting stage. The study provides predictions for intelligent production and lodging risk management.
Article
Nanoscience & Nanotechnology
Chia-Hua Hsieh, Fang-Tzu Lin, Kun-Yi Andrew Lin, Shang-Yu Hsieh, Yi-Ting Chen, Hui-Ping Tsai, Chieh-Hsuan Lu, Hongta Yang
Summary: This study successfully develops a photonic crystal material inspired by cephalopod skins, which can change its color by applying voltage. The material can maintain its appearance and lattice structure under ambient conditions and can be restored by applying an oxidation potential.
ACS APPLIED NANO MATERIALS
(2022)
Article
Agriculture, Multidisciplinary
Cheng-Ju Lee, Ming-Der Yang, Hsin-Hung Tseng, Yu-Chun Hsu, Yu Sung, Wei-Ling Chen
Summary: Single-plant growth monitoring in precision agriculture helps reduce costs and optimize decision-making. This study used UAV imagery and deep learning methods to detect and monitor individual broccoli plants, providing a visualized growth map for precise field management. The proposed approach can be applied to other crops and improve efficiency in agriculture.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Proceedings Paper
Geography, Physical
G. G. Wang, H. P. Tsai
Summary: This study examined the trends of cloud forests in Shei-Pa National Park in Taiwan and used an LSTM model to predict future vegetation status. Preliminary results showed an improvement in vegetation condition in the area, with the maximum temperature prediction model performing the best. These findings provide valuable insights for forest resource conservation and climate adaptation strategies in Taiwan.
XXIV ISPRS CONGRESS: IMAGING TODAY, FORESEEING TOMORROW, COMMISSION III
(2022)
