Model Specialization for the Use of ESRGAN on Satellite and Airborne Imagery

Clabaut, Étienne; Lemelin, M.; Germain, Mickaël; Bouroubi, Yacine; St-Pierre, Tony

doi:10.3390/rs13204044

Cited by 9 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Considering significant limitations that may artificially improve results, such as the lack of optical blur, it is uncertain whether a flight altitude of 60 m, can be definitively deemed ideal (Brown et al, 2022; Clabaut et al, 2021). The absence of optical blur might have led to unrealistically sharp degraded images due to pixel subsampling compared with actual flights at higher altitudes.…”

Section: Discussionmentioning

confidence: 99%

“…The last value corresponds to the resolution obtained when flying at 120 m, the maximum altitude permitted by Canadian regulations. Bicubic resampling method was used without the addition of optical blur, giving potentially superior results compared to real-world applications (Brown et al, 2022; Clabaut et al, 2021). The size of the tiles was then resized back to 256 x 256 pixels for all the GSD values to prevent the size of the input from impacting the model results (Richter et al, 2021).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Early Detection of an Invasive Alien Plant (Phragmites australis) Using Unoccupied Aerial Vehicles and Artificial Intelligence

Caron-Guay,

Germain,

Laliberté

2024

Preprint

Self Cite

View full text Add to dashboard Cite

The combination of unoccupied aerial vehicles (UAVs) and artificial intelligence to map vegetation represents a promising new approach to improve the detection of invasive alien plant species (IAPS). The high spatial resolution achievable with UAVs and recent innovations in computer vision, especially with convolutional neural networks, suggest that early detection of IAPS could be possible, thus facilitating their management. In this study, we evaluated the suitability of this approach for mapping the location of common reed (Phragmites australissubsp.australis) within a national park located in southern Quebec, Canada. We collected data on six distinct dates during the growing season, covering environments with different levels of reed invasion. Overall, model performance was high for the different dates and zones, especially for recall (mean of 0.89). The results showed an increase in performance, reaching a peak following the appearance of the inflorescence in September (highest F1-score at 0.98). Furthermore, a decrease in spatial resolution negatively affected recall (18% decrease between a spatial resolution of 0.15 cm pixel−1and 1.50 cm pixel−1) but did not have a strong impact on precision (2% decrease). Despite challenges associated with common reed mapping in a post-treatment monitoring context, the use of UAVs and deep learning shows great potential for IAPS detection when supported by a suitable dataset. Our results show that, from an operational point of view, this approach could be an effective tool for speeding up the work of biologists in the field and ensuring better management of IAPS.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Early Detection of an Invasive Alien Plant (Phragmites australis) Using Unoccupied Aerial Vehicles and Artificial Intelligence

Caron-Guay,

Germain,

Laliberté

2024

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…These methods are widely used in the fields of aeronautics, medicine, and engineering [14,15]. In the field of aviation, a variety of image super-resolution reconstruction algorithms had been proposed [16][17][18][19][20]. Zhou et al [18] proposed a super-resolution reconstruction strategy based on self-attention generative adversarial networks, which improves the details of remote sensing images by adding self-attention modules.…”

Section: Introductionmentioning

confidence: 99%

Manhole Cover Classification Based on Super-Resolution Reconstruction of Unmanned Aerial Vehicle Aerial Imagery

Wang,

Huang

2024

Applied Sciences

View full text Add to dashboard Cite

Urban underground pipeline networks are a key component of urban infrastructure, and a large number of older urban areas lack information about their underground pipelines. In addition, survey methods for underground pipelines are often time-consuming and labor-intensive. While the manhole cover serves as the hub connecting the underground pipe network with the ground, the generation of underground pipe network can be realized by obtaining the location and category information of the manhole cover. Therefore, this paper proposed a manhole cover detection method based on UAV aerial photography to obtain ground images, using image super-resolution reconstruction and image positioning and classification. Firstly, the urban image was obtained by UAV aerial photography, and then the YOLOv8 object detection technology was used to accurately locate the manhole cover. Next, the SRGAN network was used to perform super-resolution processing on the manhole cover text to improve the clarity of the recognition image. Finally, the clear manhole cover text image was input into the VGG16_BN network to realize the manhole cover classification. The experimental results showed that the manhole cover classification accuracy of this paper’s method reached 97.62%, which verified its effectiveness in manhole cover detection. The method significantly reduces the time and labor cost and provides a new method for manhole cover information acquisition.

show abstract

“…The work [8] is based on Super-Resolution Generative Adversarial Networks (SRGAN), where the authors modify the loss function and structure of the SRGAN network and propose an improved SRGAN (ISRGAN) that makes model training more stable and enhances the ability to generalize across locations and sensors. In the experiment, training and testing data were collected from two sensors (Landsat 8 OLI and Chinese GF 1) from different locations (Guangdong and Xinjiang in China).…”

Section: Introductionmentioning

confidence: 99%

Applying machine learning to improve a texture type image

Tussupov

Kozhabai

Bayegizova

et al. 2023

EEJET

View full text Add to dashboard Cite

The paper is devoted to machine learning methods that focus on texture-type image enhancements, namely the improvement of objects in images. The aim of the study is to develop algorithms for improving images and to determine the accuracy of the considered models for improving a given type of images. Although currently used digital imaging systems usually provide high-quality images, external factors or even system limitations can cause images in many areas of science to be of low quality and resolution. Therefore, threshold values for image processing in a certain field of science are considered. The first step in image processing is image enhancement. The issues of signal image processing remain in the focus of attention of various specialists. Currently, along with the development of information technology, the automatic improvement of images used in any field of science is one of the urgent problems. Images were analyzed as objects: state license plates of cars, faces, sections of the field on satellite images. In this work, we propose to use the models of Super-Resolution Generative Adversarial Network (SRGAN), Extended Super-Resolution Generative Adversarial Networks (ERSGAN). For this, an experiment was conducted, the purpose of which was to retrain the trained ESRGAN model with three different architectures of the convolutional neural network, i.e. VGG19, MobileNet2V, ResNet152V2 to add perceptual loss (by pixels), also add more sharpness to the prediction of the test image, and compare the performance of each retrained model. As a result of the study, the use of convolutional neural networks to improve the image showed high accuracy, that is, on average ESRGAN+MobileNETV2 – 91 %, ESRGAN+VGG19 – 86 %, ESRGAN+ResNet152V2 – 96 %.

show abstract

Model Specialization for the Use of ESRGAN on Satellite and Airborne Imagery

Cited by 9 publications

References 20 publications

Early Detection of an Invasive Alien Plant (Phragmites australis) Using Unoccupied Aerial Vehicles and Artificial Intelligence

Early Detection of an Invasive Alien Plant (Phragmites australis) Using Unoccupied Aerial Vehicles and Artificial Intelligence

Manhole Cover Classification Based on Super-Resolution Reconstruction of Unmanned Aerial Vehicle Aerial Imagery

Applying machine learning to improve a texture type image

Contact Info

Product

Resources

About