Article
Computer Science, Artificial Intelligence
Yanzhu Liu, Yanan Wang, Adams Wai Kin Kong
Summary: This paper investigates a new problem of salient object grading and proposes a pixel-wise ordinal classification method. Experimental results demonstrate that the method provides effective salient level predictions and offers comparable performance with state-of-the-art salient object detection methods in the traditional problem setting.
IMAGE AND VISION COMPUTING
(2021)
Article
Environmental Sciences
Martina Deur, Mateo Gasparovic, Ivan Balenovic
Summary: The study evaluated the influence of pansharpening of WorldView-3 satellite imagery on tree species classification, finding that the LMVM algorithm was the most effective. Object-based classification outperformed pixel-based classification, with an overall accuracy increase of 4% to 7%.
Article
Environmental Sciences
Ziyun Yan, Lei Ma, Weiqiang He, Liang Zhou, Heng Lu, Gang Liu, Guoan Huang
Summary: This study compares object-based and pixel-based LCZ mapping methods and finds that the object-based method performs better in terms of overall accuracy, especially in land cover types, and shows competitive results in certain LCZ categories. Feature selection results indicate the importance of building height, sky view factor, building surface fraction, permeable surface fraction, and land use in the object-based paradigm.
Article
Environmental Sciences
Yiyun Luo, Jinnian Wang, Xiankun Yang, Zhenyu Yu, Zixuan Tan
Summary: This study proposed a semantic segmentation method based on pixel representation augmentation, which utilizes a cross-attention mechanism in the Transformer to achieve excellent performance, effectively addressing two key issues in the transfer process from natural image segmentation to land cover classification.
Article
Computer Science, Information Systems
Lu Liu, Zixuan Xu, Daqing He, Dequan Yang, Hongchen Guo
Summary: This paper systematically studies vanishing attacks against a remote sensing image object detection model and proposes adversarial attack adaptation methods based on interpolation scaling and patch perturbation stacking. It also suggests a hot restart perturbation update strategy and a local pixel attack algorithm based on sensitive pixel location to achieve good attack effects. Experimental results show that the proposed attack method achieves a balance between attack effect and attack cost.
Article
Environmental Sciences
Soraya Yaghobi, Alireza Daneshi, Sajad Khoshnood, Hossein Azadi
Summary: This study aimed to compare the accuracy of nine different methods for identifying land use types in Malekshahi City located in Western Iran. The artificial neural network (ANN) algorithm with back-propagation algorithms achieved the highest accuracy and efficiency, with a kappa coefficient and overall accuracy of approximately 0.94 and 96.5, respectively. The methods of Mahalanobis distance (MD) and minimum distance to mean (MDM) were introduced as the next priority, with overall accuracies of about 91.35 and 90.0, respectively, for land use categorization. Considering the high accuracy, the ANN algorithm is concluded to be the best algorithm for extracting land use maps in Malekshahi City.
ENVIRONMENTAL MONITORING AND ASSESSMENT
(2023)
Article
Environmental Studies
Jun-Yi Zheng, Ying-Ying Hao, Yuan-Chen Wang, Si-Qi Zhou, Wan-Ben Wu, Qi Yuan, Yu Gao, Hai-Qiang Guo, Xing-Xing Cai, Bin Zhao
Summary: This study compared the performance of pixel-based, object-based image analysis, and deep learning methods in classifying vegetation in coastal wetlands using a dataset based on RGB-based UAV data. The results showed that the deep learning method outperformed the other methods in classifying the vegetation, accurately reflecting its distribution. The study also confirmed the ability of the object-based image analysis method on Google Earth Engine to process UAV data for the first time.
Article
Computer Science, Artificial Intelligence
Putu Desiana Wulaning Ayu, Sri Hartati, Aina Musdholifah, Detty S. Nurdiati
Summary: Amniotic fluid protects the fetus during uterine development, and a new pixel classification model is proposed to separate amniotic fluid from other objects. By testing it on amniotic fluid ultrasound images, the proposed model shows great performance compared to existing methods.
APPLIED SOFT COMPUTING
(2021)
Article
Instruments & Instrumentation
K. Schweikert, A. Sielaff, P. Stephan
Summary: This study introduces a calibration method for temperature measurement using infrared cameras, enabling high-accuracy thermography of moving targets with varying emissivity. The analysis shows that measurement uncertainty decreases with increasing sub-pixel resolution, resulting in reduced error introduced by sample motion.
INFRARED PHYSICS & TECHNOLOGY
(2021)
Article
Optics
Zhe Yang, Yu-Ming Bai, Li-Da Sun, Ke-Xin Huang, Jun Liu, Dong Ruan, Jun-Lin Li
Summary: The study proposes a new concurrent single-pixel imaging, object location, and classification scheme (SP-ILC) based on deep learning. The results demonstrate that the scheme can accurately locate, image, and classify objects with low sampling rates simultaneously, making it suitable for applications in remote sensing, medical diagnosis, security, and autonomous vehicle control.
Article
Environmental Studies
Ru Xu
Summary: This study proposed a method to map rural settlements using spectral-texture-temporal information from Landsat and Sentinel time series. By calculating spectral and texture indices and stacking them for segmentation, accurate maps of rural settlements were obtained, showing consistent accuracy in different landscape conditions.
Article
Biodiversity Conservation
Xiang Liu, Julian Frey, Martin Denter, Katarzyna Zielewska-Buettner, Nicole Still, Barbara Koch
Summary: The study introduced a pixel- and object-based image fusion method using very high-resolution stereo WV-3 data to accurately map standing dead trees. Adding canopy height model reduced errors, and combining all WV-3 derived variables achieved the highest classification accuracy. Standing dead trees were mainly distributed on higher and steeper north-facing slopes in the forest area.
ECOLOGICAL INDICATORS
(2021)
Article
Computer Science, Artificial Intelligence
Ze Song, Xudong Kang, Xiaohui Wei, Shutao Li
Summary: Camouflaged object detection is always faced with the challenge of identifying object pixels embedded in the background. Existing deep learning methods lack the ability to effectively utilize the context information around different pixels. In this paper, a pixel-centric context perception network (PCPNet) is proposed to address this problem. PCPNet customizes personalized context for each pixel based on the automatic estimation of its surroundings. Experimental results demonstrate the superiority of PCPNet in camouflaged object detection compared to other state-of-the-art methods.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Chemistry, Analytical
Abhijeet Boragule, Hyunsung Jang, Namkoo Ha, Moongu Jeon
Summary: This study introduces a pixel-guided method to efficiently build a joint detection and tracking framework for multi-object tracking. By queuing and utilizing per-pixel distributions to compute the association matrix, and introducing long-term appearance association in track features, advanced MOT performance is achieved.
Article
Optics
Manhong Yao, Shujun Zheng, Yuhang Hu, Zibang Zhang, Junzheng Peng, Jingang Zhong
Summary: Due to limited bandwidth and storage space, it is challenging to classify fast-moving objects based on high-speed photography for a long duration. In this paper, we propose a single-pixel classification method using deep learning, where the scene image is modulated by orthogonal transform basis patterns and the modulated light signal is detected by a single-pixel detector. The proposed method shows improved reliability in classification results for fast-moving objects through differential measuring and measurement data rolling utilization.
Article
Geochemistry & Geophysics
Jiang Chen, Wei-Ning Zhu, Yong Q. Tian, Qian Yu
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2017)
Article
Environmental Sciences
Jiang Chen, Weining Zhu, Yong Q. Tian, Qian Yu, Yuhan Zheng, Litong Huang
JOURNAL OF APPLIED REMOTE SENSING
(2017)
Article
Environmental Sciences
Yuhan Zheng, Carlos M. Duarte, Jiang Chen, Dan Li, Zhaohan Lou, Jiaping Wu
GEOCARTO INTERNATIONAL
(2019)
Article
Engineering, Environmental
Jiang Chen, Weining Zhu, Yuhan Zheng, Yong Q. Tian, Qian Yu
Article
Engineering, Environmental
Weining Zhu, Litong Huang, Nan Sun, Jiang Chen, Shuna Pang
WATER ENVIRONMENT RESEARCH
(2019)
Article
Environmental Sciences
Jiang Chen, Weining Zhu, Yong Q. Tian, Qian Yu
SCIENCE OF THE TOTAL ENVIRONMENT
(2020)
Article
Environmental Sciences
Jiang Chen, Tao He, Bo Jiang, Shunlin Liang
REMOTE SENSING OF ENVIRONMENT
(2020)
Article
Environmental Sciences
Niu Di, Weiliang He, Kaihua Zhang, Jian Cui, Jiang Chen, Jinming Cheng, Bao Chu, Shanshan Li, Yinyu Xie, Hao Xiang, Hebo Wang, Gongbo Chen, Yuming Guo
Summary: This study examines the short-term effects of ambient air pollution on systemic inflammatory biomarkers, revealing significant associations between ozone (O-3) and decreases in certain white blood cell types, as well as associations between PM2.5 and increases in other white blood cell types.
ENVIRONMENTAL RESEARCH LETTERS
(2021)
Article
Environmental Sciences
Yinyu Xie, Weiliang He, Xiaoling Zhang, Jian Cui, Xiaochao Tian, Jiang Chen, Kaihua Zhang, Shanshan Li, Niu Di, Hao Xiang, Hebo Wang, Gongbo Chen, Yuming Guo
Summary: This study found that long-term exposure to air pollution may be associated with the occurrence of carotid plaque, while increased greenness can reduce the risk of carotid plaque. The effects of air pollution and greenness on carotid plaque need to be further studied and addressed.
ENVIRONMENTAL POLLUTION
(2021)
Article
Construction & Building Technology
Gongbo Chen, Jiang Chen, Guang-hui Dong, Bo-yi Yang, Yisi Liu, Tianjun Lu, Pei Yu, Yuming Guo, Shanshan Li
Summary: China is facing increasing ozone pollution due to rapid economic development and urbanization. The study aimed to improve surface ozone concentration estimates using an iterative random forest model, recent ground monitoring data, and high-resolution meteorological data, achieving a higher accuracy than previous studies. The newly generated daily max 8-h average ozone data product with improved spatial resolution shows great potential in assessing the health effects of ozone pollution in both short-term and long-term studies.
SUSTAINABLE CITIES AND SOCIETY
(2021)
Article
Geochemistry & Geophysics
Jiang Chen, Tao He, Shunlin Liang
Summary: This study proposes an operational approach to estimate daily 1-km surface all-wave net radiation (R-n) using an improved hybrid model (AHM) and extended hybrid model (EHM), which combines a radiative transfer model, a machine learning algorithm, and geostationary satellite observations.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Geochemistry & Geophysics
Yichuan Ma, Tao He, Shunlin Liang, Jianguang Wen, Jean-Philippe Gastellu-Etchegorry, Jiang Chen, Anxin Ding, Siqi Feng
Summary: Surface albedo estimation is challenging in rugged terrains, but this study successfully retrieved albedo on sloping terrain using satellite data and artificial neural networks. The accuracy of the method was verified and different albedo results were evaluated, advancing our understanding of energy budget in mountains.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Geography, Physical
Weining Zhu, Zeliang Zhang, Zaiqiao Yang, Shuna Pang, Jiang Chen, Qian Cheng
Summary: This study introduces a new distribution-distribution scheme using statistical inferences to estimate the probability distribution of in-water components, providing valuable information for improving traditional methods. Analysis of Landsat-8 images showed diverse patterns in spectral probability distributions of global waters, leading to the development of a bootstrap-based scheme for estimating distribution parameters of yellow substance in water.
PHOTOGRAMMETRIC ENGINEERING AND REMOTE SENSING
(2021)