Article
Computer Science, Artificial Intelligence
Xin Tian, Rui Liu, Zhongyuan Wang, Jiayi Ma
Summary: A novel 3D reconstruction method combining polarization imaging and binocular stereo vision has been proposed, which aims to improve accuracy by correcting azimuth angle errors and utilizing low-rank matrix factorization constraints. Experimental results demonstrate the efficiency of the method and its wide application prospects in 3D reconstruction.
INFORMATION FUSION
(2022)
Article
Chemistry, Analytical
Afshin Azizi, Yousef Abbaspour-Gilandeh, Tarahom Mesri-Gundoshmian, Aitazaz A. Farooque, Hassan Afzaal
Summary: This research developed a computerized method based on stereo vision technique to estimate the roughness formed on agricultural soils, and investigated soil till quality by analyzing the height of plow layers. The results showed that the proposed method has strong potential for estimating soil shallow roughness in tillage operations, while peak fitting is an effective method for evaluating tillage quality.
Article
Chemistry, Analytical
Xinhua Wang, Dayu Li, Guang Zhang
Summary: This paper proposes an optical optimization design scheme for panoramic imaging based on binocular stereo vision, and develops a panoramic stereo real-time imaging system by combining real-time processing algorithms for multi-detector mosaic panoramic stereo imaging images. Experimental results demonstrate that the system can meet the requirements for high-definition panoramic video imaging under different lighting conditions.
Article
Optics
Lianghui Li, Jiachen Wang, Shengli Yang, Hao Gong
Summary: The study proposes a method for illuminance measurement using binocular stereo vision technology, establishing illuminance analysis models and theoretical calculation equations for the relationship between power, illuminance, and depth of LEDs. Results show that illuminance is proportional to power consumed by LEDs, and the proposed method is efficient, reliable, and in good agreement with experimental results.
Article
Computer Science, Software Engineering
Krzysztof Wolski, Fangcheng Zhong, Karol Myszkowski, Rafal K. Mantiuk
Summary: This paper introduces a model and contrast enhancement algorithm that can improve the accuracy of binocular depth cues under low brightness, allowing high-quality stereoscopic images to be displayed even at very low luminance.
ACM TRANSACTIONS ON GRAPHICS
(2022)
Article
Engineering, Multidisciplinary
Guohua Gao, Shuangyou Wang, Ciyin Shuai
Summary: This paper focuses on the tomato localization issue in the vision system of tomato picking robots. It uses a binocular camera to collect images, improves the principle of binocular ranging, and enhances the census stereo matching algorithm. The improved algorithm optimizes the area matching process, applies more constraints, and achieves an extremely small disparity error. The matching time is significantly improved, and experimental results show that the improved algorithm provides more accurate localization information.
ALEXANDRIA ENGINEERING JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Hui Wei, Lingjiang Meng
Summary: This study proposes a matching algorithm that combines segment and edge matching to improve the accuracy of stereo matching algorithms in binocular vision. By simulating the mechanism of biological vision, the algorithm transforms pixel matching to pixel segment matching and edge matching, reducing time complexity. It can be implemented in an industrial robot environment for high-precision needle threading guidance.
PATTERN RECOGNITION
(2023)
Article
Agronomy
Xiangming Lei, Mingliang Wu, Yajun Li, Anwen Liu, Zhenhui Tang, Shang Chen, Yang Xiang
Summary: In order to achieve rapid recognition and accurate picking of Camellia oleifera fruits, a binocular vision system composed of two industrial cameras was used. The YOLOv7 convolutional neural network model was trained iteratively, and the optimal weight model was selected for fruit recognition. Experimental results showed high accuracy rates, and the binocular stereo vision system demonstrated accurate depth positioning for fruit detection.
Article
Chemistry, Analytical
Ming Wei, Ming Zhu, Yi Wu, Jiaqi Sun, Jiarong Wang, Changji Liu
Summary: The paper introduces a new end-to-end fast deep learning stereo matching network that innovates in feature extraction and cost volume construction, while combining edge guidance and multi-cross attention model to achieve excellent performance in both speed and accuracy.
Article
Optics
Yuguang Hou, Changying Liu, Bowen An, Yang Liu
Summary: A stereo matching algorithm based on Census transform and texture filtering is proposed in this paper to solve the problem of low matching accuracy in discontinuous disparity and low texture area for existing binocular stereo matching algorithms. The algorithm uses the weighted Census transform circular template for matching cost calculation, which reflects the influence of the distance between neighborhood pixels and target pixels and expands the perception range of target pixels. The texture filtering method is used for cost aggregation to highlight the image structure information and smooth the internal texture. Experimental results show that the proposed algorithm effectively reduces the mismatching rate of images, produces disparity maps with less noise, and achieves better matching effect in relatively dense pattern textures.
Article
Computer Science, Artificial Intelligence
Juliano Emir Nunes Masson, Marcelo Roberto Petry, Daniel Ferreira Coutinho, Leonardo de Mello Honorio
Summary: Multi-View Stereo (MVS) is a critical step in photogrammetry, relying on the ability to match features in different images. Convolutional Neural Networks have been used to solve this problem, but they consume a large amount of Video RAM. This study reduces GPU memory usage and introduces deformable convolutions to improve the performance.
IMAGE AND VISION COMPUTING
(2022)
Article
Chemistry, Multidisciplinary
Chengtao Zhu, Yau-Zen Chang
Summary: This paper proposes a simple texture-independent aggregation approach that achieves high performance in stereo matching. The approach involves matrix multiplications of two weighting matrices and a primary matching cost to generate dense disparity maps. Additionally, a multi-scale scheme is integrated to exploit the spatial distribution of textures for higher matching accuracy. The resulting hybrid approach surpasses most existing approaches in terms of efficiency and accuracy in stereo matching.
APPLIED SCIENCES-BASEL
(2023)
Article
Chemistry, Multidisciplinary
Ting-Wei Chang, Wei-Cheng Wang, Rongshun Chen
Summary: This study introduces a novel automatic irrigation system for individual seedlings, integrating visual recognition, identification of watering points, and control of spraying nozzles. Experimental results show successful watering rates of 82% and 83.3% for uni-pot and multi-pot orchid seedlings, respectively.
APPLIED SCIENCES-BASEL
(2021)
Article
Chemistry, Analytical
Botao Liu, Kai Chen, Sheng-Lung Peng, Ming Zhao
Summary: The study introduces a novel stereo matching network, ASR-Net, which combines multi-level residual optimization and depth map super-resolution to improve accuracy and speed. Experimental results demonstrate outstanding performance on different datasets, with the three-pixel error reducing to 2.86% for the kitti2015 dataset and surpassing traditional methods in terms of speed.
Article
Computer Science, Information Systems
Jing Ding, Zhigang Yan, Xuchen We
Summary: The paper proposes a reliable and stable moving target localization method based on binocular stereo vision, which integrates various algorithms to achieve moving target recognition and extraction, improving the effectiveness, accuracy, and robustness of three-dimensional moving target coordinates.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2021)
Article
Agricultural Engineering
Hao Tian, Huirong Xu, Yibin Ying
Summary: This study investigated the penetration depth of light and the distribution of components inside pomelos to improve the accuracy of predicting soluble solids content (SSC) using visible-near infrared spectroscopy (Vis-NIRS). Three different modes of puncture measurement were conducted, and the asymmetrical changes and distribution of light intensity were observed in tissues. The semi-transmittance mode combined with the limited penetration depth and SSC distribution characteristics in pomelo was adopted. Multi-point spectra in the semi-transmittance mode were used to evaluate the influence of asymmetrical light and SSC distribution on models. The models established by mean spectra showed good performance in eliminating the difference in light distribution. The effective wavelengths were selected using competitive adaptive reweighted sampling (CARS) to establish the calibration model. The global model of CARS-PLSR using mean spectra and mean SSC of six points achieved the best performance with low root mean square error of prediction (RMSEP) and residual predictive deviation (RPD) values. Overall, multi-point detection in the semi-transmittance mode allowed non-destructive prediction of SSC in pomelo by weakening distribution difference caused by biological variability.
BIOSYSTEMS ENGINEERING
(2022)
Article
Engineering, Environmental
Qi Zhang, Zhishang Li, Huang Dai, Lin Zhang, Jie Zhang, Yuanjie Liu, Jianhan Lin, Kang Liang, Yibin Ying, Yanbin Li, Yingchun Fu
Summary: Inspired by biomineralization, the study successfully grew ultrahigh load metal-organic frameworks (MOFs) on inert glass fibers, preparing hybrid membranes with high efficiency in collecting organic hazardous substances. The new method significantly enhanced the coverage and adsorption capacity of MOFs, showing exceptional robustness and maintained performance in extreme conditions.
CHEMICAL ENGINEERING JOURNAL
(2022)
Article
Agriculture, Multidisciplinary
Jie Yang, Juntao Li, Jie Hu, Wenjun Yang, Xiaolei Zhang, Jinfan Xu, Youchao Zhang, Xuan Luo, K. C. Ting, Tao Lin, Yibin Ying
Summary: This study proposes a deep learning approach called DeepTranSpectra (DTS) for calibration transfer in spectroscopic analysis. It effectively avoids the need for standard samples by using labeled samples from slave instruments. The DTS approach demonstrates improved transfer performance compared to traditional standardization methods. Feature visualization is used to interpret the transfer mechanism of the DTS approach.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Food Science & Technology
Ruiyun Zhou, Chen Wang, Qi Wang, Lijuan Xie, Yibin Ying
Summary: An efficient method based on laser-engraved free-standing terahertz (THz) metamaterials has been developed for rapid analysis of fruit acids. The proposed strategy shows great potential for rapid in situ analysis of flavor components in food products, successfully distinguishing fruit acid concentrations as low as 0.1 mg L-1.
FOOD ANALYTICAL METHODS
(2022)
Article
Chemistry, Multidisciplinary
Tailong Cai, Lingyi Lan, Bo Peng, Chao Zhang, Shufen Dai, Chi Zhang, Jianfeng Ping, Yibin Ying
Summary: Water-enabled electricity generation technologies are attractive and renewable solutions to global energy crisis and pollution. However, current technologies still face challenges such as high cost, harmful components, and specific environmental requirements. This study presents a high-performance wood-based moisture-enabled electric generator that can offer clean energy for undeveloped and disaster-relief regions.
Article
Multidisciplinary Sciences
Yuzhou Shao, Lusong Wei, Xinyue Wu, Chengmei Jiang, Yao Yao, Bo Peng, Han Chen, Jiangtao Huangfu, Yibin Ying, Chuanfang John Zhang, Jianfeng Ping
Summary: This research presents a strategy for direct printing of flexible wireless electronics at room temperature. By regulating the additive-free titanium carbide (Ti3C2Tx) MXene aqueous inks, high-resolution and high-performance functional modules can be fabricated and integrated on various substrates. This work opens up possibilities for high-precision additive manufacturing of printed wireless electronics at room temperature.
NATURE COMMUNICATIONS
(2022)
Article
Chemistry, Multidisciplinary
Xiao Wang, Yixian Wang, Hao Qi, Yun Chen, Wei Guo, Haiyan Yu, Huayun Chen, Yibin Ying
Summary: This study developed a core-shell dye/MOFs@COFs gas-sensing material with excellent humidity resistance and sub-ppm level sensitivity, which is applicable for early and accurate detection of wheat scab.
Article
Biophysics
Lin Zhang, Yuxin Sun, Ziyan Zhang, Yafang Shen, Yue Li, Tongtong Ma, Qi Zhang, Yibin Ying, Yingchun Fu
Summary: Pesticide residues have caused public concern, and metal-organic frameworks have been used as recognition and signal generation elements in pesticide detection sensors. A sensor based on porous MOFs hybrid sponge for fluorescent-visible detection of methyl parathion has been developed, showing rapid response time, wide linear detection range, and low detection limit.
BIOSENSORS & BIOELECTRONICS
(2022)
Article
Engineering, Environmental
Huayun Chen, Zhiheng You, Xiao Wang, Qimin Qiu, Yibin Ying, Yixian Wang
Summary: Flexible metal-organic framework (MOF) sensor allows high selectivity detection of VOCs and exhibits excellent reversibility and storage stability. It also shows potential in identifying plant diseases caused by pathogen infection.
CHEMICAL ENGINEERING JOURNAL
(2022)
Article
Chemistry, Physical
Tongtong Ma, Jie Zhang, Lin Zhang, Qi Zhang, Xiahong Xu, Yonghua Xiong, Yibin Ying, Yingchun Fu
Summary: This paper summarizes the progress of emerging films based on nanomaterials and their detection applications in recent five years, focusing on typical electrochemical and optical methods. Some new interesting applications, such as point-of-care testing, wearable devices and detection chips, are proposed and emphasized. This review will provide insights into the integration and processability of films based on nanomaterials, thus stimulate further contributions towards films based on nanomaterials for high-performance analytical-chemistry-related applications.
ADVANCES IN COLLOID AND INTERFACE SCIENCE
(2023)
Article
Agriculture, Multidisciplinary
Dihua Wu, Yibin Ying, Mingchuan Zhou, Jinming Pan, Di Cui
Summary: A chicken gender identification method based on an improved ResNet-50 deep learning algorithm was proposed in this study. The algorithm achieved better overall performance than five typical recognition algorithms and four state-of-the-art (SOTA) animal gender identification methods, with an accuracy rate of 98.42%, precision of 97.92%, recall of 98.95%, F1 score of 98.43%, and inference time of 4.79 ms.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Chemistry, Multidisciplinary
Lin Zhang, Yuxin Sun, Li Peng, Wenzhang Fang, Qiao Huang, Jie Zhang, Ziyan Zhang, Hang Li, Yingjun Liu, Yibin Ying, Yingchun Fu
Summary: This study developed a dynamic bridging strategy based on fibrin to fabricate hierarchical porous materials with multi-scale pores. The fabricated material has a large surface area and high adsorption capacity.
Review
Materials Science, Multidisciplinary
Shufen Dai, Xunjia Li, Chengmei Jiang, Jianfeng Ping, Yibin Ying
Summary: Triboelectric nanogenerator (TENG) efficiently harvests mechanical energy from the agricultural environment to provide distributed power for microdevice networks in smart agriculture, enabling self-powered agricultural monitoring and production strategy adjustment and further promoting agricultural production intelligence.
Article
Engineering, Electrical & Electronic
Chao Zhang, Chi Zhang, Xinyue Wu, Jianfeng Ping, Yibin Ying
Summary: The integrated plant wearable system (IPWS) based on adaptive winding strain (AWS) sensor can provide stable and high-fidelity monitoring of plant pulse, reflecting the growth and water state of plants in real time.
NPJ FLEXIBLE ELECTRONICS
(2022)
Article
Food Science & Technology
Xunjia Li, Jianjun Luo, Kai Han, Xue Shi, Zewei Ren, Yi Xi, Yibin Ying, Jianfeng Ping, Zhong Lin Wang
Summary: This article introduces a wind and rain energy-driven electrical stimulation system for enhancing crop production. The system utilizes a triboelectric nanogenerator and a raindrop-driven nanogenerator, which can significantly increase the germination speed and yield of pea seeds through the self-generated high-voltage electric field. By harnessing environmental energy, the system can drive agricultural sensors to optimize plant growth.