PSGCNet: A Pyramidal Scale and Global Context Guided Network for Dense Object Counting in Remote-Sensing Images

Gao, Guangshuai; Liu, Qingjie; Hu, Zhenghui; Lü, Li; Wen, Quan; Wang, Yunhong

doi:10.1109/tgrs.2022.3153946

Cited by 20 publications

(4 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To provide comprehensive benchmarks, we evaluate some popular object-counting methods on the NWPU-MOC dataset. These methods include the crowd-counting method (MCNN [24], CSRNet [39], SFCN [52], SCAR [53]) and the objectcounting methods designed for remote sensing scenes (ASP-Net [14], PSGCNet [15], SwinCounter [50]), respectively. The experimental results on the NWPU-MOC dataset, where two types of Gaussian kernels are used to generate Gaussian density maps, are presented in Table II.…”

Section: B Mainstream Methods Involved In Evaluationmentioning

confidence: 99%

“…Gao et al [14] construct an RSOC dataset for remote sensing object counting, containing four subsets: buildings, small vehicles, large vehicles, and boats. Based on the dataset, Gao et al propose the PSGCNet [15], which addresses challenges such as scale variations and complex backgrounds in remote sensing scenes by extracting and fusing multi-scale and global feature information.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Feature Balance for Fine-Grained Object Classification in Aerial Images

Zhao

Tong

Yao

et al. 2022

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Object counting is a hot topic in computer vision, which aims to estimate the number of objects in a given image. However, most methods only count objects of a single category for an image, which cannot be applied to scenes that need to count objects with multiple categories simultaneously, especially in aerial scenes. To this end, this paper introduces a Multi-category Object Counting (MOC) task to estimate the numbers of different objects (cars, buildings, ships, etc.) in an aerial image. Considering the absence of a dataset for this task, a large-scale Dataset (NWPU-MOC) is collected, consisting of 3,416 scenes with a resolution of 1024 × 1024 pixels, and well-annotated using 14 fine-grained object categories. Besides, each scene contains RGB and Near Infrared (NIR) images, of which the NIR spectrum can provide richer characterization information compared with only the RGB spectrum. Based on NWPU-MOC, the paper presents a multi-spectrum, multicategory object counting framework, which employs a dualattention module to fuse the features of RGB and NIR and subsequently regress multi-channel density maps corresponding to each object category. In addition, to modeling the dependency between different channels in the density map with each object category, a spatial contrast loss is designed as a penalty for overlapping predictions at the same spatial position. Experimental results demonstrate that the proposed method achieves state-ofthe-art performance compared with some mainstream counting algorithms. The dataset, code and models are publicly available at https://github.com/lyongo/NWPU-MOC.

show abstract

Section: B Mainstream Methods Involved In Evaluationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Feature Balance for Fine-Grained Object Classification in Aerial Images

Zhao

Tong

Yao

et al. 2022

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Wan et al [26] propose an end-to-end framework to learn the density map of the counter and achieve good performance. Gao et al [27] propose a new remote sensing object counting network PSGCNet, which includes a Pyramid Scale Module (PSM) and a Global Context Module (GCM).…”

Section: Remote Sensing Object Countingmentioning

confidence: 99%

Global Multi-Scale Information Fusion for Multi-Class Object Counting in Remote Sensing Images

Gong

2022

Remote Sensing

View full text Add to dashboard Cite

In recent years, object counting has been investigated and has made significant progress under a surveillance-view. However, there exists only a few works focusing on the remote sensing object density estimation, and the performance of existing methods is not promising. On the one hand, due to the imbalance distribution of targets in remote sensing images, the model might collapse, leading a severe degradation. On the other hand, the scale of targets in remote sensing images actually varies in real scenarios, which remains a challenge for counting objects accurately. To remedy the above problems, we propose an approach named “SwinCounter” for object counting in remote sensing. Moreover, we introduce a Balanced MSE Loss to pay more attention to the fewer samples, which alleviates the problem of imbalanced object labels. In addition, the attention mechanism in our SwinCounter can precisely capture multi-scale information. Thus, the model is aware of different scales of objects, which capture small and dense targetes more precisely. We build experiments on the RSOC dataset, achieving MAEs of 7.2, 151.5, 14.38, and 52.88 and MSEs of 10.1, 436.0, 22.7, and 74.82 on the Building, Small-Vehicle, Large-Vehicle, and Ship sub-datasets, which demonstrates the competitiveness and superiority of the proposed method.

show abstract

“…Nevertheless, several variables including scale fluctuations, climate changes, perception abnormalities, and orientation modifications have an impact on the effectiveness of these old methods [13,14]. Deep learning identification techniques including single shot multibox detectors (SSD) and regional-convolutional neural networks (R-CNN) have newly reached great effectiveness and proposed a possible answer to these difficulties [15]. In spite of the popularity of deep learning technologies, a standardized collection of olive trees is not accessible for deep learning purposes.…”

Section: Introductionmentioning

confidence: 99%

An Efficient Deep Learning Mechanism for the Recognition of Olive Trees in Jouf Region

Alshammari

Shahin²

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

Olive trees grow all over the world in reasonably moderate and dry climates, making them fortunate and medicinal. Pesticides are required to improve crop quality and productivity. Olive trees have had important cultural and economic significance since the early pre-Roman era. In 2019, Al-Jouf region in a Kingdom of Saudi Arabia’s north achieved global prominence by breaking a Guinness World Record for having more number of olive trees in a world. Unmanned aerial systems (UAS) were increasingly being used in aerial sensing activities. However, sensing data must be processed further before it can be used. This processing necessitates a huge amount of computational power as well as the time until transmission. Accurately measuring the biovolume of trees is an initial step in monitoring their effectiveness in olive output and health. To overcome these issues, we initially formed a large scale of olive database for deep learning technology and applications. The collection comprises 250 RGB photos captured throughout Al-Jouf, KSA. This paper employs among the greatest efficient deep learning occurrence segmentation techniques (Mask Regional-CNN) with photos from unmanned aerial vehicles (UAVs) to calculate the biovolume of single olive trees. Then, using satellite imagery, we present an actual deep learning method (SwinTU-net) for identifying and counting of olive trees. SwinTU-net is a U-net-like network that includes encoding, decoding, and skipping links. SwinTU-net’s essential unit for learning locally and globally semantic features is the Swin Transformer blocks. Then, we tested the method on photos with several wavelength channels (red, greenish, blues, and infrared region) and vegetation indexes (NDVI and GNDVI). The effectiveness of RGB images is evaluated at the two spatial rulings: 3 cm /pixel and 13 cm /pixel, whereas NDVI and GNDV images have only been evaluated at 13 cm/pixel. As a result of integrating all datasets of GNDVI and NDVI, all generated mask regional-CNN-based systems performed well in segmenting tree crowns (F1-measure from 95.0 to 98.0 percent). Based on ground truth readings in a group of trees, a calculated biovolume was 82 percent accurate. These findings support all usage of NDVI and GNDVI spectrum indices in UAV pictures to accurately estimate the biovolume of distributed trees including olive trees.

show abstract

PSGCNet: A Pyramidal Scale and Global Context Guided Network for Dense Object Counting in Remote-Sensing Images

Cited by 20 publications

References 71 publications

Feature Balance for Fine-Grained Object Classification in Aerial Images

Feature Balance for Fine-Grained Object Classification in Aerial Images

Global Multi-Scale Information Fusion for Multi-Class Object Counting in Remote Sensing Images

An Efficient Deep Learning Mechanism for the Recognition of Olive Trees in Jouf Region

Contact Info

Product

Resources

About