Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Chun-Ta Wei has a total of 39 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Environment aging analysis of animal bloodstains with ATR-FTIR and CNN

    by Chun-Ta Wei, Zexin Shen, Wenbin Luo, Jingyi Zhao, Tingting Yin, Kaining Cheng, Miao Zhang

  • Open Access

    Low-Frequency Active Noise Control System Based on Feedback FXLMS

    by Ting-Yu Chen, Jia-Horng Yang, Chien-Liang Lai, Chun-Ta Wei

    The rise of industrial machinery and military training activities has significantly contributed to low-frequency noise pollution, which can penetrate traditional passive noise isolation methods and pose serious health risks, including irreversible hearing damage. To address this challenge, this study proposes a hybrid active noise control (HANC) system, integrating an adaptive active noise control (AANC) module based on the filtered-x least mean squares (FxLMS) algorithm and an audio-balance control circuit (ABCC). The AANC system actively generates anti-noise signals to mitigate low-frequency disturbances, while the ABCC module enhances voice clarity and protects users from excessive impulse noise. MATLAB R2023b simulations and hardware implementations validate the system’s effectiveness, achieving a noise reduction of up to 21.8 dB in controlled environments. Additionally, the proposed feedback active noise control architecture ensures robust performance under dynamic noise conditions, improving stability and response time. By integrating both software-based adaptive filtering and hardware circuit design, this study provides a comprehensive noise mitigation solution with potential applications in military, industrial, and vehicular environments.


  • Open Access

    Research on Soft Error Detection Method for Self-Similar Data Streams Based on Structure Mapping

    by Junjie Wu, Songjian Huang, Senpeng Chen, Chun-Ta Wei, Jiangqiang Hu, Kaiming Ji, Irfan Khan, Miao Zhang

    Self-similar data streams are characterized by their similarity across multiple time scales, exhibiting distinct nonlinear and discrete features. These characteristics complicate the accurate identification of data points associated with soft error features, thereby making it difficult to effectively discern the intricate relationship between the data flow and soft error data. This, in turn, severely impacts the accuracy of soft error detection in self-similar data streams. In this study, we propose a novel approach to address this challenge. Leveraging the suddenness and long-range correlation inherent in self-similar data streams, we construct a time series model to capture the data stream based on linear correlation and straight-line fitting features. By incorporating relationship parameters to fit neighboring flow points, we utilize a structural mapping model to establish the local angular relationship between the data streams and soft error data. Additionally, we construct a structural mapping network using flow features to achieve soft error detection in self-similar data streams. Experimental results demonstrate that our proposed method achieves high accuracy and low time overhead for soft error detection in self-similar data streams.

  • Using Deep Neural Networks Analyzing Wound Images for Forensic Tool Mark Identification

    by Chun-Ta Wei, Jeff Cheng-Lung Lee, Yu-Guang Chen, Wei-Hung Lin, Hui-Tzu Ting, Chi-Tse Pai

    The crime scene often reveals various types of wounds caused by weapons. It is the primary task of the first responders at the scene to determine the type of weapon based on the wound patterns. For newly graduated or inexperienced personnel, developing a mobile device capable of instantly identifying wounds and corresponding weapons would significantly enhance the efficiency and accuracy of crime scene investigations. This study utilizes deep neural networks to analyze the features of wound images and deploys the neural network on a portable Field Programmable Gate Array (FPGA) for hardware acceleration and offline recognition. Due to limited real wound photo data, we initially use simulated wounds made of resin clay as training data. By combining image processing techniques, we designed a neural network model with high accuracy and resolution. Experimental results demonstrate that this device has advantages such as lightweight, fast operation, accuracy, and offline capability, effectively overcoming environmental variables at the scene and providing objective analysis results to aid in the identification of weapons in crime scene investigations.


  • Open Access

    A court line extraction algorithm for badminton tournament videos with horizontal line projection learning

    by Chun‐Ta Wei, Shiuh‐Ku Weng

    Court line extraction is one of the important steps in the analysis of sport videos. The court extraction is the foundation of the analysis of badminton video, and an efficient method with horizontal line projection K‐means machine learning algorithm to extract court lines from different broadcast badminton tournament videos is proposed in this paper. The horizontal lines are projected into 1‐D histogram signal; then the signal is trained to learn the intensity of the histogram signal for locating the positions of the horizontal court lines. After the equations of the horizontal court lines and the court lines in the vertical direction have been formulized, the intersection points of the court lines can be calculated and the court line can be extracted. The experimental results show that the proposed method can extract the court lines more efficiently than that done by the Hough transform‐related algorithms, which are widely applied in computer vision and self‐driving car applications.


  • Open Access

    Exploring Airborne LiDAR and Aerial Photographs Using Machine Learning for Land Cover Classification

    by Ming-Da Tsai, Kuan-Wen Tseng, Chia-Cheng Lai, Chun-Ta Wei, Ken-Fa Cheng

    Airborne LiDAR is a popular measurement technology in recent years. Its feature is that it can quickly acquire high precision and high density 3D point coordinates on the surface. The reflective waveform of the radar contains the geometric structure and roughness of the surface reflector. Combined with the information from aerial photographs, it can quickly help users to interpret various surface object types and serve as a basis for land cover classification. The experiment is divided into three phases. In the phase 1, LiDAR data and decision tree classification method (DT) were used to classify the land cover and customize the geometric parameter elevation. In the phase 2, we combined aerial photographs, LiDAR data and DT method to improve the accuracy of land cover classification. In the phase 3, the support vector machine classification method (SVM) was used to compare the classification accuracy of different classification methods. The results show that customizing the geometric parameter elevation can improve the overall classification accuracy. The results of the study showed that the DT method and the SVM method had better results for the grass, building and artificial ground, and the SVM method had better results for the planted shrub and bare ground.


  • Open Access

    Enhancing the Accuracy of Land Cover Classification by Airborne LiDAR Data and WorldView-2 Satellite Imagery

    by Chun-Ta Wei, Ming-Da Tsai, Yu-Lung Chang, Ming-Chih Jason Wang

    The Full Waveform LiDAR system has been developed and used commercially all over the world. It acts to record the complete time of a laser pulse and has a high-resolution sampling interval compared to the traditional multiple-echo LiDAR, which only provides signals within a single target range. This study area mainly collects data from Riegl LMS-Q680i Full Waveform LiDAR and WorldView-2 satellite imagery, which focuses on buildings, vegetation, grassland, asphalt roads and other ground types as the surface objects. The amplitude and pulse width are selected as waveform basic parameters. The parameter of topography is slope, and the height classification parameters of the test ground are 0–0.5 m, 0.5–2.5 m, and 2.5 m. To eliminate noise, the neighborhood average is applied on the LiDAR parameter values and analyzed as the classification accuracy comparison. This survey uses Decision Tree as the classification method. Comparing the data between neighborhood average and non-neighborhood average, the data classification accuracy improves by 7%, and Kappa improves by 5.92%. NDVI image data are utilized to distinguish the artificial from natural ground. The results show that the neighborhood average with previous data can improve the classification accuracy by 5%, and Kappa improves by 4.25%. By adding NIR-2 of WorldView-2 satellite imagery to the neighborhood average analysis, the overall classification accuracy is improved by 2%, and the Kappa value by 1.21%. This article shows that utilizing the analysis of neighborhood average and image parameters can effectively improve the classification accuracy of land covers.