Road extraction from aerial images has been a hot research topic in the field of remote sensing image analysis. In this letter, a semantic segmentation neural network, which combines the strengths of residual learning and U-Net, is proposed for road area extraction. The network is built with residual units and has similar architecture to that of U-Net. The benefits of this model are twofold: first, residual units ease training of deep networks. Second, the rich skip connections within the network could facilitate information propagation, allowing us to design networks with fewer parameters, however, better performance. We test our network on a public road data set and compare it with U-Net and other two state-of-the-art deep-learning-based road extraction methods. The proposed approach outperforms all the comparing methods, which demonstrates its superiority over recently developed state of the arts.

Road Extraction by Deep Residual U-Net

Deep Convolutional Neural Networks have been adopted for salient object detection and achieved the state-of-the-art performance. Most of the previous works however focus on region accuracy but not on the boundary quality. In this paper, we propose a predict-refine architecture, BASNet, and a new hybrid loss for Boundary-Aware Salient object detection. Specifically, the architecture is composed of a densely supervised Encoder-Decoder network and a residual refinement module, which are respectively in charge of saliency prediction and saliency map refinement. The hybrid loss guides the network to learn the transformation between the input image and the ground truth in a three-level hierarchy -- pixel-, patch- and map- level -- by fusing Binary Cross Entropy (BCE), Structural SIMilarity (SSIM) and Intersection-over-Union (IoU) losses. Equipped with the hybrid loss, the proposed predict-refine architecture is able to effectively segment the salient object regions and accurately predict the fine structures with clear boundaries. Experimental results on six public datasets show that our method outperforms the state-of-the-art methods both in terms of regional and boundary evaluation measures. Our method runs at over 25 fps on a single GPU. The code is available at: https://github.com/NathanUA/BASNet.

/pdf/basnet-boundary-aware-salient-object-detection-13wvhaxy88.pdf

BASNet: Boundary-Aware Salient Object Detection

Image segmentation is a key topic in image processing and computer vision with applications such as scene understanding, medical image analysis, robotic perception, video surveillance, augmented reality, and image compression, among many others. Various algorithms for image segmentation have been developed in the literature. Recently, due to the success of deep learning models in a wide range of vision applications, there has been a substantial amount of works aimed at developing image segmentation approaches using deep learning models. In this survey, we provide a comprehensive review of the literature at the time of this writing, covering a broad spectrum of pioneering works for semantic and instance-level segmentation, including fully convolutional pixel-labeling networks, encoder-decoder architectures, multi-scale and pyramid based approaches, recurrent networks, visual attention models, and generative models in adversarial settings. We investigate the similarity, strengths and challenges of these deep learning models, examine the most widely used datasets, report performances, and discuss promising future research directions in this area.

Image Segmentation Using Deep Learning: A Survey

Deep learning (DL) based semantic segmentation methods have been providing state-of-the-art performance in the last few years. More specifically, these techniques have been successfully applied to medical image classification, segmentation, and detection tasks. One deep learning technique, U-Net, has become one of the most popular for these applications. In this paper, we propose a Recurrent Convolutional Neural Network (RCNN) based on U-Net as well as a Recurrent Residual Convolutional Neural Network (RRCNN) based on U-Net models, which are named RU-Net and R2U-Net respectively. The proposed models utilize the power of U-Net, Residual Network, as well as RCNN. There are several advantages of these proposed architectures for segmentation tasks. First, a residual unit helps when training deep architecture. Second, feature accumulation with recurrent residual convolutional layers ensures better feature representation for segmentation tasks. Third, it allows us to design better U-Net architecture with same number of network parameters with better performance for medical image segmentation. The proposed models are tested on three benchmark datasets such as blood vessel segmentation in retina images, skin cancer segmentation, and lung lesion segmentation. The experimental results show superior performance on segmentation tasks compared to equivalent models including U-Net and residual U-Net (ResU-Net).

Recurrent Residual Convolutional Neural Network based on U-Net (R2U-Net) for Medical Image Segmentation.

During the outbreak time of COVID-19, computed tomography (CT) is a useful manner for diagnosing COVID-19 patients. Due to privacy issues, publicly available COVID-19 CT datasets are highly difficult to obtain, which hinders the research and development of AI-powered diagnosis methods of COVID-19 based on CTs. To address this issue, we build an open-sourced dataset -- COVID-CT, which contains 349 COVID-19 CT images from 216 patients and 463 non-COVID-19 CTs. The utility of this dataset is confirmed by a senior radiologist who has been diagnosing and treating COVID-19 patients since the outbreak of this pandemic. We also perform experimental studies which further demonstrate that this dataset is useful for developing AI-based diagnosis models of COVID-19. Using this dataset, we develop diagnosis methods based on multi-task learning and self-supervised learning, that achieve an F1 of 0.90, an AUC of 0.98, and an accuracy of 0.89. According to the senior radiologist, models with such performance are good enough for clinical usage. The data and code are available at this https URL

COVID-CT-Dataset: A CT Scan Dataset about COVID-19

Automatic extraction of ground objects is fundamental for many applications of remote sensing. It is valuable to extract different kinds of ground objects effectively by using a general method. We propose such a method, JointNet, which is a novel neural network to meet extraction requirements for both roads and buildings. The proposed method makes three contributions to road and building extraction: (1) in addition to the accurate extraction of small objects, it can extract large objects with a wide receptive field. By switching the loss function, the network can effectively extract multi-type ground objects, from road centerlines to large-scale buildings. (2) This network module combines the dense connectivity with the atrous convolution layers, maintaining the efficiency of the dense connection connectivity pattern and reaching a large receptive field. (3) The proposed method utilizes the focal loss function to improve road extraction. The proposed method is designed to be effective on both road and building extraction tasks. Experimental results on three datasets verified the effectiveness of JointNet in information extraction of road and building objects.

JointNet: A Common Neural Network for Road and Building Extraction

Urban functional zones refer to areas (or regions) of a city which provide specific urban functions for peoples who lived in the city. The spatial layout of buildings in functional zone show a specific pattern, e.g. residual areas usually have similar builds and the positions of which are highly organized. In this paper, we show that it is possible to identify urban functional zones from a remote sensed imagery. To this end, a convolutional neural networks (CNN) based functional zone classification method is proposed. The method mainly consists of three steps. Firstly, the aerial imagery of the city is partitioned into disjoint regions by road network. Then, each region is further divided into patches and is fed to a fully connected CNN. The output of which is considered as distributions of this patches on the five previously defined functional zones. Finally, we take a vote strategy to identify the function zone of this region. We test our method on a collection of Google Earth images over Shenyang, Beijing, etc. The results demonstrate the effectiveness of the proposed method.

A CNN based functional zone classification method for aerial images

The invention discloses a functional region dividing method and device based on a remote-sensing image, and relates to the technical field of artificial intelligence. The method comprises that an SIFT (Scale-invariant feature transform) feature is extracted from the remote-sensing image to be identified; according to the SIFT feature of the remote-sensing image to be identified as well as an image feature expression dictionary, an image feature histogram of the remote-sensing image to be identified is generated; and according to the image feature histogram and a functional region classifier, a functional region of the remote-sensing image to be identified is determined. According to the invention, city functional areas of different satellite data can be divided automatically, rapidly, conveniently and accurately, and automatic functional region division has substantial application values in the aspects including city planning, disaster prevention, emergent disaster rescue and business investigation.

Functional region dividing method and device based on remote-sensing image

In recent years, UAV remote sensing has gradually attracted the attention of scientific researchers and industry, due to its broad application prospects. It has been widely used in agriculture, forestry, mining, and other industries. UAVs can be flexibly equipped with various sensors, such as optical, infrared, and LIDAR, and become an essential remote sensing observation platform. Based on UAV remote sensing, researchers can obtain many high-resolution images, with each pixel being a centimeter or millimeter. The purpose of this paper is to investigate the current applications of UAV remote sensing, as well as the aircraft platforms, data types, and elements used in each application category; the data processing methods, etc.; and to study the advantages of the current application of UAV remote sensing technology, the limitations, and promising directions that still lack applications. By reviewing the papers published in this field in recent years, we found that the current application research of UAV remote sensing research can be classified into four categories according to the application field: (1) Precision agriculture, including crop disease observation, crop yield estimation, and crop environmental observation; (2) Forestry remote sensing, including forest disease identification, forest disaster observation, etc.; (3) Remote sensing of power systems; (4) Artificial facilities and the natural environment. We found that in the papers published in recent years, image data (RGB, multi-spectral, hyper-spectral) processing mainly used neural network methods; in crop disease monitoring, multi-spectral data are the most studied type of data; for LIDAR data, current applications still lack an end-to-end neural network processing method; this review examines UAV platforms, sensors, and data processing methods, and according to the development process of certain application fields and current implementation limitations, some predictions are made about possible future development directions.

https://www.mdpi.com/2504-446X/7/6/398/pdf?version=1686822643

Zhengxin Zhang

Papers

Road Extraction by Deep Residual U-Net

JointNet: A Common Neural Network for Road and Building Extraction

A CNN based functional zone classification method for aerial images

Functional region dividing method and device based on remote-sensing image

A Review on Unmanned Aerial Vehicle Remote Sensing: Platforms, Sensors, Data Processing Methods, and Applications