Article
Spectroscopy
Yong Hao, Yuanhang Lu, Xiyan Li
Summary: In this study, a stability monitor model was established using the multivariate statistical process control (MSPC) method, and a mixed modeling approach combining robust regression (Rob-Reg) and partial least squares regression (PLSR) was employed to eliminate the variability influence of sample and instrument states. The results showed that MSPC effectively monitored the consistency of the same batch samples measured at different times or different batches, and the Rob-Reg method outperformed the PLSR method in predicting the different batches of samples.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2022)
Article
Environmental Sciences
Weichao Liu, Hongyuan Huo, Ping Zhou, Mingyue Li, Yuzhen Wang
Summary: This paper proposes a new method for predicting soil total iron composition. The method screens abnormal samples using a Monte Carlo method based on particle swarm optimization, and adopts feature representation based on Shannon entropy for wavelet packet processing. The selected feature bands based on the correlation coefficient and the CARS algorithm are applied to the soil spectra before and after wavelet packet processing. Finally, a 1D-CNN is used to calculate the Fe content. Experimental results show that the method can effectively handle abnormal samples, improve the correlation between spectra and content, and achieve good results with few samples.
Article
Spectroscopy
Xiao-Wen Zhang, Zheng-Guang Chen, Feng Jiao
Summary: The dimensionality of near-infrared (NIR) spectral data is often large, and dimensionality reduction is crucial for increasing the model's performance. Laplacian Eigenmaps (LE) can preserve local neighborhood information but is disturbed by irrelevant information and multicollinearity. Random Frog (RF) algorithm can eliminate noise and collinearity. Hence, before using LE, RF is used to eliminate irrelevant information and reduce correlation, resulting in improved regression models' prediction accuracy and stability.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2023)
Article
Chemistry, Analytical
Jiani Li, Fanfan Liang, Li Han, Xiaoxuan Yu, Dingbin Liu, Wensheng Cai, Xueguang Shao
Summary: In this study, the characteristic absorption of water in the near-infrared (NIR) region was utilized to determine extra-cellular pH (pHe) and intra-cellular pH (pHi). The use of continuous wavelet transform (CWT) enhanced the resolution of the spectra and partial least squares (PLS) regression was employed to establish quantitative models for pHe and pHi. The results showed that water with different hydrogen bonds can serve as a good probe to sense pH within biological systems.
Article
Spectroscopy
Yi Lu, Xiaolong Li, Weijiao Li, Tingting Shen, Zhenni He, Mengqi Zhang, Hao Zhang, Yongqi Sun, Fei Liu
Summary: The study successfully quantified chlorpyrifos and carbendazim residues in cabbage using visible/near-infrared spectroscopy combined with chemometric methods. Preprocessing and feature variable selection were employed, with LS-SVM models performing well on global spectra data and SPA-selected feature variables showing good performance in carbendazim detection.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2021)
Article
Environmental Sciences
J. Antonio Q. Guzman, G. Arturo Sanchez-Azofeifa
Summary: Using wavelet spectra can improve the accuracy of predicting leaf traits over reflectance models and require fewer components. Additionally, the effect of plant groups on model performance is significant, with traits of lianas generally showing lower R2 compared to tree traits.
REMOTE SENSING OF ENVIRONMENT
(2021)
Article
Agricultural Engineering
Jinming Liu, Shuo Jin, Changhao Bao, Yong Sun, Wenzhe Li
Summary: This study proposed a rapid detection method based on near-infrared reflectance spectroscopy to measure the contents of cellulose, hemicellulose, and lignin in corn stover. By constructing a BiPLS-PCA-SVM model, the study demonstrated an alternative strategy for detecting lignocellulosic components in pre-treated corn stover in the anaerobic digestion process.
BIORESOURCE TECHNOLOGY
(2021)
Article
Chemistry, Analytical
Haoran Li, Suyi Chen, Jisheng Dai, Xiaobo Zou, Tao Chen, Tianhong Pan, Melvin Holmes
Summary: The proposed method introduces a new fast burst-sparsity learning approach for baseline correction, utilizing downsampling strategy and pattern-coupled prior to overcome the limitations of existing baseline correction methods. The study demonstrates that burst-sparsity commonly occurs in peak zones of spectra and can be properly utilized to enhance baseline correction performance.
ANALYTICAL CHEMISTRY
(2022)
Article
Automation & Control Systems
Abbas Saadatmandi, Mahmoud Reza Sohrabi, Hasan Kabiri Fard
Summary: In this study, a fast, easy, inexpensive, and precise method combining UV-Vis spectrophotometry, continuous wavelet transform, and partial least squares multivariate calibration was developed for the simultaneous determination of paracetamol, diphenhydramine, and phenylephrine in tablet dosage form without extraction. The method was validated using synthetic mixtures and showed good performance.
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2023)
Article
Biotechnology & Applied Microbiology
Huazhou Chen, Hanli Qiao, Quanxi Feng, Lili Xu, Qinyong Lin, Ken Cai
Summary: Advanced chemometric methods were investigated for the detection of pomelo fruit quality using near-infrared hyperspectral imaging (NIRHI) technology. The proposed RBF-PLS model was optimized for parameter scaling and showed good predictive accuracy for sugar, vitamin C, and organic acid content in pomelo samples. The combination of NIRHI technology and chemometric methods is applicable for rapid quantitative detection of pomelo fruit quality, with potential for detection in other agricultural products.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2021)
Article
Spectroscopy
Gerard Dumancas, Indra Adrianto
Summary: This study developed a stacked regression ensemble approach using near infrared spectroscopic method for accurate determination of biomass compositional analyses. The performance of various machine learning techniques was compared, and the stacked regression outperformed other methods, providing a more accurate prediction of biomass compositions.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2022)
Article
Chemistry, Multidisciplinary
Huaichun Xiao, Yang Liu, Yande Liu, Hui Xiao, Liwei Sun, Yong Hao
Summary: This study investigated the feasibility of using visible and near-infrared spectra for rapid detection and improved identification accuracy of citrus greening through spectral fusion. The results showed that a least squares support vector machine model based on principal component analysis had the best performance in feature-level fusion. The accuracy of this model was 100% for the detection of citrus greening.
APPLIED SCIENCES-BASEL
(2023)
Article
Green & Sustainable Science & Technology
So-Yeon Jeong, Eun-Ju Lee, Se-Eun Ban, Jae-Won Lee
Summary: Near infrared (NIR) spectroscopy is a rapid, accurate, and non-destructive method for analyzing biomass composition, but prediction accuracy varies for different biomass compositions. Different biomass particle sizes have statistically significant differences in NIR spectra based on root mean square values, and preprocessing methods do not significantly improve prediction accuracy.
Article
Food Science & Technology
Lei-Ming Yuan, Lifan You, Xiaofeng Yang, Xiaojing Chen, Guangzao Huang, Xi Chen, Wen Shi, Yiye Sun
Summary: A strategy of fusing consensus models based on the genetic algorithm was proposed to measure the soluble solids content in peaches. The consensus models achieved better performance in predicting the SSC of peaches compared to traditional optimized models.
Article
Chemistry, Analytical
Yong-Qi Zhong, Jia-Qi Li, Xiao-Long Li, Sheng-Yun Dai, Fei Sun
Summary: This study aimed to develop a rapid, simple, and accurate method using near-infrared spectroscopy coupled with chemometrics to quantify five main chemical components in Arnebiae Radix. The results demonstrated that near-infrared spectroscopy could be a promising approach for the rapid quality assessment of Arnebiae Radix.
VIBRATIONAL SPECTROSCOPY
(2023)
Review
Food Science & Technology
Yong He, Qinlin Xiao, Xiulin Bai, Lei Zhou, Fei Liu, Chu Zhang
Summary: Fruits are susceptible to damage during their growth, harvest, and storage, which can impact both food safety and economic benefits. To address this issue, there is a need for rapid and nondestructive detection methods for fruit damage. This paper summarizes various nondestructive techniques for detecting fruit damage, providing insights for future research and real-world applications.
CRITICAL REVIEWS IN FOOD SCIENCE AND NUTRITION
(2022)
Article
Agriculture, Multidisciplinary
Yufei Liu, Jichun Wang, Yachao Shi, Zhenni He, Fei Liu, Wenwen Kong, Yong He
Summary: The rapid development of new technologies such as automatic control, sensors, and AI has greatly contributed to the advancement of unmanned airboats (UA) and their applications in fields like environmental monitoring and agriculture. This review presents the challenges and potential solutions in the development of UA, along with its structure, comprehensive applications, and future prospects. It provides theoretical and technical support for the promotion of UA for automated operations.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Computer Science, Information Systems
Xiangyu Lu, Rui Yang, Jun Zhou, Jie Jiao, Fei Liu, Yufei Liu, Baofeng Su, Peiwen Gu
Summary: This study proposes an effective and accurate approach based on Ghost-convolution and Transformer networks for diagnosing grape leaf in the field. The results show that the proposed method achieves high accuracy and fast processing speed, making it suitable for diagnosing grape diseases and pests in vineyards.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Biochemistry & Molecular Biology
Mahamed Lamine Guindo, Muhammad Hilal Kabir, Rongqin Chen, Jing Huang, Fei Liu, Xiaolong Li, Hui Fang
Summary: This study investigated the combination of LIBS and Vis-NIR for fast detection of phosphorus (P) and potassium (K) in organic fertilizers. XAI was used to extract valuable features from both sensors, and the fusion of data was performed. The outcomes showed that the fusion method was more efficient in detecting P and K compared to single-sensor detection.
Article
Environmental Sciences
Jiyu Peng, Yifan Liu, Longfei Ye, Jiandong Jiang, Fei Zhou, Fei Liu, Jing Huang
Summary: Minerals in rice leaves are important indicators of plant health and are used to guide plant management. This study used LIBS to predict mineral content in rice leaves under Cr stress. PLSR achieved good performance in predicting Ca, Fe, Mg, K, Mn, and Na concentrations. The correlation between different spectral lines was also analyzed. This method provides a fast and accurate approach for predicting minerals in rice leaves under Cr stress, which is crucial for environmental protection and food safety.
SCIENCE OF THE TOTAL ENVIRONMENT
(2023)
Article
Remote Sensing
Xiangyu Lu, Jun Zhou, Rui Yang, Zhiyan Yan, Yiyuan Lin, Jie Jiao, Fei Liu
Summary: This study presents a novel approach to extract and map phenological traits of rice directly from unmanned aerial vehicle (UAV) photographs. A multi-stage rice field segmentation dataset named PaddySeg was built, and an efficient Ghost Bilateral Network (GBiNet) was proposed to generate trait masks. The mapping of rice phenology was achieved by interpolation on trait value-location pairs.
Article
Engineering, Environmental
Wei Wang, Zun Man, Xiaolong Li, Rongqin Chen, Zhengkai You, Tiantian Pan, Xiaorong Dai, Hang Xiao, Fei Liu
Summary: This study investigated the effect of cadmium on root phenotypes by examining cadmium accumulation, adversity physiology, morphological parameters, and microstructure characteristics. The study found that cadmium had both a low-promotion and high-inhibition effect on root phenotypes. Additionally, rapid detection methods for cadmium accumulation and adversity physiology were developed using spectroscopic technology and chemometrics, significantly reducing the detection time compared to laboratory analysis.
JOURNAL OF HAZARDOUS MATERIALS
(2023)
Article
Engineering, Environmental
Xiaolong Li, Jing Huang, Rongqin Chen, Zhengkai You, Jiyu Peng, Qingcai Shi, Gang Li, Fei Liu
Summary: Rapid and accurate detection of agricultural soil chromium is crucial for soil pollution assessment. Laser-induced breakdown spectroscopy (LIBS) is a rapid and chemical-free method for hazardous metal analysis, but its detection is interfered by uncertainty and matrix effect. In this study, a strategy combining linear weighted network (LWNet) was proposed to reduce uncertainty, and the AWN-LWNet framework was proposed to reduce the matrix effect in two soil types. The results indicated that LWNet outperformed traditional machine learning and achieved an average relative error of 2.08% and 3.03% for yellow brown soil and lateritic red soil, respectively. AWN-LWNet was the optimal model to reduce matrix effect (ARE=4.12%). Besides, AWN-LWNet greatly reduced the number...
JOURNAL OF HAZARDOUS MATERIALS
(2023)
Article
Biochemistry & Molecular Biology
Zhengkai You, Xiaolong Li, Jing Huang, Rongqin Chen, Jiyu Peng, Wenwen Kong, Fei Liu
Summary: In this study, a novel liquid-solid conversion method based on agarose films was proposed for the detection of heavy metals in water. The method involved converting water samples into semi-solid hydrogels using agarose and drying them into agarose films to enhance the signal intensities. Calibration curves were constructed for Cd, Pb, and Cr, and the method was validated using standard heavy metal solutions and real water samples. The results demonstrated the effectiveness of the method, with high values of R-2 and low LOD values, as well as good recovery rates. The agarose film-based liquid-solid conversion method shows great potential for heavy metal monitoring in water.
Article
Biochemistry & Molecular Biology
Xinmeng Luo, Rongqin Chen, Muhammad Hilal Kabir, Fei Liu, Zhengyu Tao, Lijuan Liu, Wenwen Kong
Summary: In this study, laser-induced breakdown spectroscopy (LIBS) was used to detect the heavy metal content in Fritillaria thunbergii. Quantitative prediction models were established using a back-propagation neural network (BPNN) optimized using the particle swarm optimization (PSO) algorithm and sparrow search algorithm (SSA). The results showed that the optimized BPNN models had better accuracy than the unoptimized model. The SSA-BP model had the advantage of faster speed and higher prediction accuracy at low concentrations.
Article
Food Science & Technology
Muhammad Hilal Kabir, Mahamed Lamine Guindo, Rongqin Chen, Xinmeng Luo, Wenwen Kong, Fei Liu
Summary: This study used laser-induced breakdown spectroscopy coupled with variable selection and chemometrics to quickly and accurately detect heavy metals (Cd, Cu, and Pb) in Fritillaria thunbergii by analyzing selected variables. The results showed that this method can improve detection efficiency and accuracy.
Proceedings Paper
Computer Science, Artificial Intelligence
Fei Liu, Fengxu Zhou, Fei Zhang, Wujing Cao
Summary: This study proposes an innovative fall detection and alarm system for the elderly in the family environment based on deep learning. The system utilizes a camera and an edge device to detect and alert users to falls without touching their body. With the use of a lightweight object detection model and an inference engine, the system achieves high accuracy and comfort in fall detection.
INTELLIGENT ROBOTICS AND APPLICATIONS (ICIRA 2022), PT I
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Fei Liu, Wujing Cao, Qingmei Li
Summary: This study proposes a fast and efficient detection method based on image processing to improve the efficiency of packaging box sorting and reduce labor intensity. It involves pose estimation and transformation relationship solving from camera coordinate system to manipulator's base coordinate system. SIFT method is used for packaging feature points, FLANN method for matching, and EPnP method for pose solving. The nine-point calibration method is used to solve the transformation relationship. The test results show that the method achieves satisfactory results by weighing detection accuracy and speed.
INTELLIGENT ROBOTICS AND APPLICATIONS (ICIRA 2022), PT I
(2022)
Article
Environmental Sciences
Yuchao Zhu, Jun Zhou, Yinhui Yang, Lijuan Liu, Fei Liu, Wenwen Kong
Summary: The study proposes an improved YOLOv4 model, which combines Mobilenetv3 network, CBAM module, and ASFF module, and optimizes the detection and counting of fruit tree canopies using the K-means algorithm, linear scaling, and cosine annealing learning strategy. The results show that the improved model can achieve fast and accurate recognition and counting of fruit tree canopies in orchard environments, with high detection accuracy and counting precision.
Article
Remote Sensing
Jun Zhou, Xiangyu Lu, Rui Yang, Huizhe Chen, Yaliang Wang, Yuping Zhang, Jing Huang, Fei Liu
Summary: This study develops a novel yield index by fusing multiple indices based on unmanned aerial vehicle imagery, which shows great potential in crop yield monitoring.