Urban Change Detection for Multispectral Earth Observation Using Convolutional Neural Networks

Daudt, Rodrigo Caye; Saux, B. Le; Boulch, Alexandre; Gousseau, Yann

doi:10.1109/igarss.2018.8518015

Cited by 304 publications

(254 citation statements)

References 8 publications

Supporting

Mentioning

251

Contrasting

Unclassified

Order By: Relevance

“…A critical question is how to define a changed patch and an unchanged patch. Some previous patch-based classification work assigned the label of the central pixel of a patch to the whole patch [44,45]. However, this definition method is sensitive to slight displacements of the patch location.…”

Section: Methodsmentioning

confidence: 99%

Detecting Building Changes between Airborne Laser Scanning and Photogrammetric Data

et al. 2019

View full text Add to dashboard Cite

Detecting topographic changes in an urban environment and keeping city-level point clouds up-to-date are important tasks for urban planning and monitoring. In practice, remote sensing data are often available only in different modalities for two epochs. Change detection between airborne laser scanning data and photogrammetric data is challenging due to the multi-modality of the input data and dense matching errors. This paper proposes a method to detect building changes between multimodal acquisitions. The multimodal inputs are converted and fed into a light-weighted pseudo-Siamese convolutional neural network (PSI-CNN) for change detection. Different network configurations and fusion strategies are compared. Our experiments on a large urban data set demonstrate the effectiveness of the proposed method. Our change map achieves a recall rate of 86.17%, a precision rate of 68.16%, and an F1-score of 76.13%. The comparison between Siamese architecture and feed-forward architecture brings many interesting findings and suggestions to the design of networks for multimodal data processing.

show abstract

Section: Methodsmentioning

confidence: 99%

Detecting Building Changes between Airborne Laser Scanning and Photogrammetric Data

et al. 2019

View full text Add to dashboard Cite

show abstract

“…A critical question is how to define a changed patch and an unchanged patch. Some previous patch-based classification work assigned the label of the central pixel of a patch to the whole patch (Hu and Yuan 2016;Daudt et al, 2018). However, this definition method is sensitive to slight displacement of the patch.…”

Section: Pre-processingmentioning

confidence: 99%

Change Detection Between Digital Surface Models From Airborne Laser Scanning and Dense Image Matching Using Convolutional Neural Networks

Zhang

Vosselman

Gerke³

et al. 2019

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

<p><strong>Abstract.</strong> Airborne photogrammetry and airborne laser scanning are two commonly used technologies used for topographical data acquisition at the city level. Change detection between airborne laser scanning data and photogrammetric data is challenging since the two point clouds show different characteristics. After comparing the two types of point clouds, this paper proposes a feed-forward Convolutional Neural Network (CNN) to detect building changes between them. The motivation from an application point of view is that the multimodal point clouds might be available for different epochs. Our method contains three steps: First, the point clouds and orthoimages are converted to raster images. Second, square patches are cropped from raster images and then fed into CNN for change detection. Finally, the original change map is post-processed with a simple connected component analysis. Experimental results show that the patch-based recall rate reaches 0.8146 and the precision rate reaches 0.7632. Object-based evaluation shows that 74 out of 86 building changes are correctly detected.</p>

show abstract

“…For example, R.C. Daudt et al [15] published an urban change-detection method based on convolutional neural networks [16] and Siamese networks [17] using bi-temporal multispectral images. Although the difference image was not used, the detection result was still not very accurate, especially at the construction boundaries.…”

Section: Introductionmentioning

confidence: 99%

CORN: An Alternative Way to Utilize Time-Series Data of SAR Images in Newly Built Construction Detection

et al. 2020

View full text Add to dashboard Cite

The limitations in obtaining sufficient datasets for training deep learning networks is preventing many applications from achieving accurate results, especially when detecting new constructions using time-series satellite imagery, since this requires at least two images of the same scene and it must contain new constructions in it. To tackle this problem, we introduce Chronological Order Reverse Network (CORN)-an architecture for detecting newly built constructions in time-series SAR images that does not require a large quantity of training data. The network uses two U-net adaptations to learn the changes between images from both Time 1-Time 2 and Time 2-Time 1 formats, which allows it to learn double the amount of changes in different perspectives. We trained the network with 2028 pairs of 256 × 256 pixel SAR images from ALOS-PALSAR, totaling 4056 pairs for the network to learn from, since it learns from both Time 1-Time 2 and Time 2-Time 1. As a result, the network can detect new constructions more accurately, especially at the building boundary, compared to the original U-net trained by the same amount of training data. The experiment also shows that the model trained with CORN can be used with images from Sentinel-1. The source code is available at https://github.com/Raveerat-titech/CORN. Remote Sens. 2020, 12, 990 2 of 25 Remote Sens. 2020, 12, 990 3 of 25 the same ground truth data. While normally, the detection of new buildings is supposed to use the data in Time 1-Time 2 format, our proposed architecture takes both Time 1-Time 2 and Time 2-Time 1 formats of data to allow learning based on both of the changing features to make it more viable. With this architecture, the amount of training data of the network appears to be doubled. This allows the network to be trained with a greater variation of data, and can result an increased detection accuracy without having to use more SAR data or create any additional ground truths. Moreover, CORN has the potential to use SAR images from other satellites and other environments because the training back and forth causes the model to be more robust.In summary, the objective of this paper is to cope with the lack of training data when training deep learning networks, which leads to inaccurate results in newly built construction detection. We do so by proposing a network architecture called "CORN", which doubles the training set by reversing the chronological order of the dataset. CORN contains two U-net adaptations; one trains on Time 1-Time 2 images, and the other one trains on Time 2-Time 1 images. The proposed network not only increases the detection accuracy, but can also be used in a greater variety of settings of the data; specifically, images from other satellites and other acquisition conditions, including the terrain of the testing area.

show abstract

Urban Change Detection for Multispectral Earth Observation Using Convolutional Neural Networks

Cited by 304 publications

References 8 publications

Detecting Building Changes between Airborne Laser Scanning and Photogrammetric Data

Detecting Building Changes between Airborne Laser Scanning and Photogrammetric Data

Change Detection Between Digital Surface Models From Airborne Laser Scanning and Dense Image Matching Using Convolutional Neural Networks

CORN: An Alternative Way to Utilize Time-Series Data of SAR Images in Newly Built Construction Detection

Contact Info

Product

Resources

About