Yun Zhang

Yun Zhang

Auto-generated profile, based on publicly available information

Yun Zhang

Department of Geodesy and Geomatics Engineering, University of New Brunswick, Fredericton, Canada, Canada

Research Focus & Metrics

Academic Career Length

24Years (2003-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 July 2018 in Remote Sensing
Verified
Journal Article
Open access
PDF file
Masoud MahdianpariMasoud Mahdianpari
Masoud Mahdianpari
Bahram SalehiBahram Salehi
Bahram Salehi
MRMohammad Rezaee
Mohammad Rezaee
FMFariba Mohammadimanesh
Fariba Mohammadimanesh
YZYun Zhang
Yun Zhang
Despite recent advances of deep Convolutional Neural Networks (CNNs) in various computer vision tasks, their potential for classification of multispectral remote sensing images has not been thoroughly explored. In particular, the applications of deep CNNs using optical remote sensing data have focused on the classification of very high-resolution aerial and satellite data, owing to the similarity of these data to the large datasets in computer vision. Accordingly, this study presents a detailed investigation of state-of-the-art deep learning tools for classification of complex wetland classes using multispectral RapidEye optical imagery. Specifically, we examine the capacity of seven well-known deep convnets, namely DenseNet121, InceptionV3, VGG16, VGG19, Xception, ResNet50, and InceptionResNetV2, for wetland mapping in Canada. In addition, the classification results obtained from deep CNNs are compared with those based on conventional machine learning tools, including Random Forest and Support Vector Machine, to further evaluate the efficiency of the former to classify wetlands. The results illustrate that the full-training of convnets using five spectral bands outperforms the other strategies for all convnets. InceptionResNetV2, ResNet50, and Xception are distinguished as the top three convnets, providing state-of-the-art classification accuracies of 96.17%, 94.81%, and 93.57%, respectively. The classification accuracies obtained using Support Vector Machine (SVM) and Random Forest (RF) are 74.89% and 76.08%, respectively, considerably inferior relative to CNNs. Importantly, InceptionResNetV2 is consistently found to be superior compared to all other convnets, suggesting the integration of Inception and ResNet modules is an efficient architecture for classifying complex remote sensing scenes such as wetlands.

2 July 2018 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Closed access
MRMohammad Rezaee
Mohammad Rezaee
Masoud MahdianpariMasoud Mahdianpari
Masoud Mahdianpari
YZYun Zhang
Yun Zhang
Bahram SalehiBahram Salehi
Bahram Salehi
The synergistic use of spatial features with spectral properties of satellite images enhances thematic land cover information, which is of great significance for complex land cover mapping. Incorporating spatial features within the classification scheme have been mainly carried out by applying just low-level features, which have shown improvement in the classification result. By contrast, the application of high-level spatial features for classification of satellite imagery has been underrepresented. This study aims to address the lack of high-level features by proposing a classification framework based on convolutional neural network (CNN) to learn deep spatial features for wetland mapping using optical remote sensing data. Designing a fully trained new convolutional network is infeasible due to the limited amount of training data in most remote sensing studies. Thus, we applied fine tuning of a pre-existing CNN. Specifically, AlexNet was used for this purpose. The classification results obtained by the deep CNN were compared with those based on well-known ensemble classifiers, namely random forest (RF), to evaluate the efficiency of CNN. Experimental results demonstrated that CNN was superior to RF for complex wetland mapping even by incorporating the small number of input features (i.e., three features) for CNN compared to RF (i.e., eight features). The proposed classification scheme is the first attempt, investigating the potential of fine-tuning pre-existing CNN, for land cover mapping. It also serves as a baseline framework to facilitate further scientific research using the latest state-of-art machine learning tools for processing remote sensing data.

25 February 2019 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
QLQingpeng Li
Qingpeng Li
LMLichao Mou
Lichao Mou
QXQizhi Xu
Qizhi Xu
YZYun Zhang
Yun Zhang
XZXiao Zhu
Xiao Zhu
Vehicle detection is a significant and challenging task in aerial remote sensing applications. Most existing methods detect vehicles with regular rectangle boxes and fail to offer the orientation of vehicles. However, the orientation information is crucial for several practical applications, such as the trajectory and motion estimation of vehicles. In this paper, we propose a novel deep network, called a rotatable region-based residual network (R³-Net), to detect multioriented vehicles in aerial images and videos. More specially, R³-Net is utilized to generate rotatable rectangular target boxes in a half coordinate system. First, we use a rotatable region proposal network (R-RPN) to generate rotatable region of interests (R-RoIs) from feature maps produced by a deep convolutional neural network. Here, a proposed batch averaging rotatable anchor strategy is applied to initialize the shape of vehicle candidates. Next, we propose a rotatable detection network (R-DN) for the final classification and regression of the R-RoIs. In R-DN, a novel rotatable position-sensitive pooling is designed to keep the position and orientation information simultaneously while downsampling the feature maps of R-RoIs. In our model, R-RPN and R-DN can be trained jointly. We test our network on two open vehicle detection image data sets, namely, DLR 3K Munich Data set and VEDAI Data set, demonstrating the high precision and robustness of our method. In addition, further experiments on aerial videos show the good generalization capability of the proposed method and its potential for vehicle tracking in aerial videos. The demo video is available at https://youtu.be/xCYD-tYudN0.

3 August 2012 in Remote Sensing
Verified
Journal Article
Open access
PDF file
Bahram SalehiBahram Salehi
Bahram Salehi
YZYun Zhang
Yun Zhang
MZMing Zhong
Ming Zhong
VDVivek Dey
Vivek Dey
Land cover classification of very high resolution (VHR) imagery over urban areas is an extremely challenging task. Impervious land covers such as buildings, roads, and parking lots are spectrally too similar to be separated using only the spectral information of VHR imagery. Additional information, therefore, is required for separating such land covers by the classifier. One source of additional information is the vector data, which are available in archives for many urban areas. Further, the object-based approach provides a more effective way to incorporate vector data into the classification process as the misregistration between different layers is less problematic in object-based compared to pixel-based image analysis. In this research, a hierarchical rule-based object-based classification framework was developed based on a small subset of QuickBird (QB) imagery coupled with a layer of height points called Spot Height (SH) to classify a complex urban environment. In the rule-set, different spectral, morphological, contextual, class-related, and thematic layer features were employed. To assess the general applicability of the rule-set, the same classification framework and a similar one using slightly different thresholds applied to larger subsets of QB and IKONOS (IK), respectively. Results show an overall accuracy of 92% and 86% and a Kappa coefficient of 0.88 and 0.80 for the QB and IK Test image, respectively. The average producers’ accuracies for impervious land cover types were also 82% and 74.5% for QB and IK.

22 November 2013 in International Journal of Image and Data Fusion
Verified
Journal Article
Closed access
YZYun Zhang
Yun Zhang
RMRakesh Kumar Mishra
Rakesh Kumar Mishra
The pan-sharpening algorithm, known as UNB PanSharp, was adopted by PCI Geomatics in 2002 and by DigitalGlobe (DG) in 2003, resulting in PCI PanSharp and DG PanSharp. Now, UNB PanSharp is developed as a stand-alone software tool, named Fuze Go™, that can be used stand-alone, integrated into ENVI and potentially into other major geo-spatial software packages such as EDARS, ESRI, and SOCET. In addition, it is in the process to integrate Fuze Go into GeoMarketSpace – a new online system that turns all geospatial information into answers. What has made UNB PanSharp successful for the last 10 years and still standing out from millions of research publications in the area? This article will go through the general principles of individual pan-sharpening algorithms that have been adopted by industry and widely used by users globally to find out the differences between UNB PanSharp and the other pan-sharpening algorithms.

1 February 2025 in Remote Sensing of Environment
Verified
Journal Article
Open access
FTFei Tong
Fei Tong
YZYun Zhang
Yun Zhang