Fusion of deep learning with adaptive bilateral filter for building outline extraction from remote sensing imagery

Khoshboresh-Masouleh, Mehdi; Shah-Hosseini, Reza

doi:10.1117/1.jrs.12.046018

Cited by 22 publications

(16 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The role of an appropriate loss function in deep convolutional neural networks is of great importance [ 48 – 50 ]. Because of the large amount of training data in deep learning, many loss functions have poor performance because of the incorrect (even low) data.…”

Section: Proposed Methodsmentioning

confidence: 99%

A Deep Learning Method for Near-Real-Time Cloud and Cloud Shadow Segmentation from Gaofen-1 Images

Khoshboresh-Masouleh

Shah-Hosseini

2020

Computational Intelligence and Neuroscience

Self Cite

View full text Add to dashboard Cite

In this study, an essential application of remote sensing using deep learning functionality is presented. Gaofen-1 satellite mission, developed by the China National Space Administration (CNSA) for the civilian high-definition Earth observation satellite program, provides near-real-time observations for geographical mapping, environment surveying, and climate change monitoring. Cloud and cloud shadow segmentation are a crucial element to enable automatic near-real-time processing of Gaofen-1 images, and therefore, their performances must be accurately validated. In this paper, a robust multiscale segmentation method based on deep learning is proposed to improve the efficiency and effectiveness of cloud and cloud shadow segmentation from Gaofen-1 images. The proposed method first implements feature map based on the spectral-spatial features from residual convolutional layers and the cloud/cloud shadow footprints extraction based on a novel loss function to generate the final footprints. The experimental results using Gaofen-1 images demonstrate the more reasonable accuracy and efficient computational cost achievement of the proposed method compared to the cloud and cloud shadow segmentation performance of two existing state-of-the-art methods.

show abstract

Section: Proposed Methodsmentioning

confidence: 99%

A Deep Learning Method for Near-Real-Time Cloud and Cloud Shadow Segmentation from Gaofen-1 Images

Khoshboresh-Masouleh

Shah-Hosseini

2020

Computational Intelligence and Neuroscience

Self Cite

View full text Add to dashboard Cite

show abstract

“…In this study, two semantic segmentation methods based on DL architectures, including Res-U-Net [24], and ABF+SegNet [16], have been used for BFE results comparisons. The models have been selected because of their good performance in BFE from multi-sensor data.…”

Section: Methodsmentioning

confidence: 99%

“…In the past few years, DL approaches play a crucial role in analysing the big image data (Khoshboresh Masouleh and Shah-Hosseini, 2019a;Maggiori et al, 2017;Masouleh and Shah-Hosseini, 2018;Samuel R. et al, 2019). The DL approaches incorporate two influential concepts in an optimal big data analysis workflow for image data .…”

Section: Deep Learningmentioning

confidence: 99%

“…Highlight of characteristics ConvNet+ SignedDist (Yuan, 2018) -Integrating multi-layer information and a unique output representation -Combine signed-distance labels with ConvNet ABF+SegNet (Masouleh and Shah-Hosseini, 2018) -Fusion of convolutional layer with adaptive bilateral filter -The minimum bounding rectangle were used for outline regularization SegNet-Dist-Fused (Yang et al, 2018) -Combine signed-distance labels with SegNet -Encoding stage based on residual learning network -Improving feature learning with a gated feature labelling unit Table 1. Overview of recent BFE researches using DL models…”

Section: Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

Robust Building Footprint Extraction From Big Multi-Sensor Data Using Deep Competition Network

Khoshboresh-Masouleh¹,

Saradjian

2019

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Commission VI, WG VI/4 ABSTRACT:Building footprint extraction (BFE) from multi-sensor data such as optical images and light detection and ranging (LiDAR) point clouds is widely used in various fields of remote sensing applications. However, it is still challenging research topic due to relatively inefficient building extraction techniques from variety of complex scenes in multi-sensor data. In this study, we develop and evaluate a deep competition network (DCN) that fuses very high spatial resolution optical remote sensing images with LiDAR data for robust BFE. DCN is a deep superpixelwise convolutional encoder-decoder architecture using the encoder vector quantization with classified structure. DCN consists of five encoding-decoding blocks with convolutional weights for robust binary representation (superpixel) learning. DCN is trained and tested in a big multi-sensor dataset obtained from the state of Indiana in the United States with multiple building scenes. Comparison results of the accuracy assessment showed that DCN has competitive BFE performance in comparison with other deep semantic binary segmentation architectures. Therefore, we conclude that the proposed model is a suitable solution to the robust BFE from big multi-sensor data.

show abstract

“…Support vector machine (SVM) has obvious advantages in solving small sample, nonlinear and high-dimensional pattern recognition, and avoids dimension disaster to some exten [19]- [21]. It can classify both linear learning and nonlinear problem via the kernel fnction [22], [23]. Generally, RSI pixels often contain a variety of earth object information, the random distribution and confusion is serious (i.e., linear and nonlinear features coexist in RSI pixels), and the classification are more complicated than that of traitional digital image.…”

Section: Introductionmentioning

confidence: 99%

Classification for Remote Sensing Data With Improved CNN-SVM Method

Sun

Liu

et al. 2019

IEEE Access

View full text Add to dashboard Cite

The efficient classification of remote sensing images (RSIs) has become the key of remote sensing application. To tackle the high computational cost in the traditional classification method, in this paper we propose a new RSI classification method based on improved convolutional neural network (CNN) and support vector machine (SVM) (CNN-SVM). In this method, we first designed a seven-layer CNN structure and took the ReLU function as the activation function. We then inputted the RSI into the CNN model and extracted feature maps and replaced the output layer of the CNN network via training the feature maps in the SVM classifier. Next, taking the simulation experiments of MNIST handwritten digital dataset and UC Merced Land Use remote sensing dataset as examples, we tested and verified the proposed method in this experiment. Finally, the empirical study of volcanic ash cloud (VAC) classification from moderate resolution imaging spectroradiometer (MODIS) RSI was carried out and evaluated. The experimental results show that compared with the traditional methods, the proposed method has lower loss value and better generalization in modeling training; the total classification accuracy of VAC and Kappa coefficient reached 93.5% and 0.8502, respectively, and achieved preferable VAC identification and visual effects. It will enhance the classification accuracy to the massive remote sensing data.

show abstract

Fusion of deep learning with adaptive bilateral filter for building outline extraction from remote sensing imagery

Cited by 22 publications

References 42 publications

A Deep Learning Method for Near-Real-Time Cloud and Cloud Shadow Segmentation from Gaofen-1 Images

A Deep Learning Method for Near-Real-Time Cloud and Cloud Shadow Segmentation from Gaofen-1 Images

Robust Building Footprint Extraction From Big Multi-Sensor Data Using Deep Competition Network

Classification for Remote Sensing Data With Improved CNN-SVM Method

Contact Info

Product

Resources

About