RAANet: A Residual ASPP with Attention Framework for Semantic Segmentation of High-Resolution Remote Sensing Images

Liu, Runrui; Tao, Fei; Liu, Xintao; Na, Jiaming; Leng, Hongjun; Wu, Junjie; Zhou, Tong

doi:10.3390/rs14133109

Cited by 68 publications

(44 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, many researchers have addressed the characteristics of remote-sensing images by proposing a high-resolution method [17][18][19]. With deep learning models, it is currently possible to segment fire pixels and determine the exact shape of a flame or smoke from various aerial images.…”

Section: Uav-based Fire Segmentation Methodsmentioning

confidence: 99%

Deep Encoder–Decoder Network-Based Wildfire Segmentation Using Drone Images in Real-Time

2022

View full text Add to dashboard Cite

Wildfire is a hazardous natural phenomenon that leads to significant human fatalities, catastrophic environmental damages, and economic losses. Over the past few years, the intensity and frequency of fires have increased worldwide. Studies have been conducted to develop distinctive solutions to minimize forest fires. Systems for distant fire detection and monitoring have been established, showing improvements in data collection and fire characterization. However, wildfires cover vast areas, making other proposed ground systems unsuitable for optimal coverage. Unmanned aerial vehicles (UAVs) have become the subject of active research in recent years. Deep learning-based image-processing methods demonstrate improved performance in various tasks, including detection and segmentation, which can be utilized to develop modern forest firefighting techniques. In this study, we established a novel two-pathway encoder–decoder-based model to detect and accurately segment wildfires and smoke from the images captured using UAVs in real-time. Our proposed nested decoder uses pre-activated residual blocks and an attention-gating mechanism, thereby improving segmentation accuracy. Moreover, to facilitate robust and generalized training, we prepared a new dataset comprising actual incidences of forest fires and smoke, varying from small to large areas. In terms of practicality, the experimental results reveal that our method significantly outperforms existing detection and segmentation methods, despite being lightweight. In addition, the proposed model is reliable and robust for detecting and segmenting drone camera images from different viewpoints in the presence of wildfire and smoke.

show abstract

Section: Uav-based Fire Segmentation Methodsmentioning

confidence: 99%

Deep Encoder–Decoder Network-Based Wildfire Segmentation Using Drone Images in Real-Time

2022

View full text Add to dashboard Cite

show abstract

“…The Deeplabv3 network 23 as a whole is divided into Encoder and Decoder parts. The encoder part mainly includes the backbone and ASPP 24 parts. In the Decoder part, the low-level features from the backbone middle layer and the high-level features output from the ASPP module are received as inputs.…”

Section: B Deeplabv3 Network and Its Improvementmentioning

confidence: 99%

X-ray Void Defect Detection in Chip Solder Joints Based on PCB-DeepLabV3 Algorithm

Kong,

Hu,

Gong

et al. 2023

Preprint

View full text Add to dashboard Cite

Defects within chip solder joints are usually inspected visually for defects using X-ray imaging to obtain images. The phenomenon of voids inside solder joints is one of the most likely types of defects in the soldering process, and accurate detection of voids becomes difficult due to their irregular shapes, varying sizes, and defocused edges. To address this problem, an X-ray void image segmentation algorithm based on improved PCB-DeepLabV3 is proposed. Firstly, to meet the demand for lightweight and easy deployment in industrial scenarios, mobilenetv2 is used as the feature extraction backbone network of the PCB-DeepLabV3 model; then, Attention Multiscale Spatial Double Pyramid Network (AMTPNet) is designed to optimize the shallow feature edges and to improve the ability to capture detailed information; finally, image cropping and cleaning methods are designed to enhance the training dataset, and the improved PCB-DeepLabV3 is applied to the training dataset. The improved PCB-DeepLabV3 model is used to segment the void regions within the solder joints and compared with the classical semantic segmentation models such as Unet, SegNet, PSPNet, and DeeplabV3. The proposed new method enables the solder joint void inspection to get rid of the traditional way of visual inspection, realize intelligent upgrading, and effectively improve the problem of difficult segmentation of the target virtual edges, to obtain the inspection results with higher accuracy.

show abstract

“…Many other models have used attention mechanisms that are popular today, being applied to many computer vision tasks such as semantic segmentation [55,56] or object detection [57]. Chen et al [58] introduced the learning of weights for multi-scale features trained with images of different sizes.…”

Section: Semantic Segmentationmentioning

confidence: 99%

SEG-ESRGAN: A Multi-Task Network for Super-Resolution and Semantic Segmentation of Remote Sensing Images

2022

View full text Add to dashboard Cite

The production of highly accurate land cover maps is one of the primary challenges in remote sensing, which depends on the spatial resolution of the input images. Sometimes, high-resolution imagery is not available or is too expensive to cover large areas or to perform multitemporal analysis. In this context, we propose a multi-task network to take advantage of the freely available Sentinel-2 imagery to produce a super-resolution image, with a scaling factor of 5, and the corresponding high-resolution land cover map. Our proposal, named SEG-ESRGAN, consists of two branches: the super-resolution branch, that produces Sentinel-2 multispectral images at 2 m resolution, and an encoder–decoder architecture for the semantic segmentation branch, that generates the enhanced land cover map. From the super-resolution branch, several skip connections are retrieved and concatenated with features from the different stages of the encoder part of the segmentation branch, promoting the flow of meaningful information to boost the accuracy in the segmentation task. Our model is trained with a multi-loss approach using a novel dataset to train and test the super-resolution stage, which is developed from Sentinel-2 and WorldView-2 image pairs. In addition, we generated a dataset with ground-truth labels for the segmentation task. To assess the super-resolution improvement, the PSNR, SSIM, ERGAS, and SAM metrics were considered, while to measure the classification performance, we used the IoU, confusion matrix and the F1-score. Experimental results demonstrate that the SEG-ESRGAN model outperforms different full segmentation and dual network models (U-Net, DeepLabV3+, HRNet and Dual_DeepLab), allowing the generation of high-resolution land cover maps in challenging scenarios using Sentinel-2 10 m bands.

show abstract

RAANet: A Residual ASPP with Attention Framework for Semantic Segmentation of High-Resolution Remote Sensing Images

Cited by 68 publications

References 30 publications

Deep Encoder–Decoder Network-Based Wildfire Segmentation Using Drone Images in Real-Time

Deep Encoder–Decoder Network-Based Wildfire Segmentation Using Drone Images in Real-Time

X-ray Void Defect Detection in Chip Solder Joints Based on PCB-DeepLabV3 Algorithm

SEG-ESRGAN: A Multi-Task Network for Super-Resolution and Semantic Segmentation of Remote Sensing Images

Contact Info

Product

Resources

About