Abstract

Airborne laser scanning (ALS) point cloud data are suitable for digital terrain model (DTM) extraction given its high accuracy in elevation. Existing filtering algorithms that eliminate non-ground points mostly depend on terrain feature assumptions or representations; these assumptions result in errors when the scene is complex. This paper proposes a new method for ground point extraction based on deep learning using deep convolutional neural networks (CNN). For every point with spatial context, the neighboring points within a window are extracted and transformed into an image. Then, the classification of a point can be treated as the classification of an image; the point-to-image transformation is carefully crafted by considering the height information in the neighborhood area. After being trained on approximately 17 million labeled ALS points, the deep CNN model can learn how a human operator recognizes a point as a ground point or not. The model performs better than typical existing algorithms in terms of error rate, indicating the significant potential of deep-learning-based methods in feature extraction from a point cloud.

Highlights

  • In recent decades, airborne laser scanning (ALS) has become more important in the process of digital terrain model (DTM) production [1]

  • After being trained on approximately 17 million labeled ALS points, the deep convolutional neural networks (CNN) model can learn how a human operator recognizes a point as a ground point or not

  • The major steps include the calculation of context information for each point from the neighboring points in a window, the transformation of the information of the window into an image, and the training and classification based on the images using the CNN model

Read more

Summary

Introduction

Airborne laser scanning (ALS) has become more important in the process of digital terrain model (DTM) production [1]. These methods maintain a surface model of the ground based on the interpolation of ground points [6,7,8,9] These methods are sensitive to input parameters and negative outliers. The existing methods have done well in ALS filtration, they still need much human labor to generate DTM based on the filtration results. ALS and responding DTM by learning a deep neural network from a big amount of the existing data. We propose a new filtering algorithm based on deep. A deep CNN model is trained using the labeled data. The deep CNN model can learn the important feature of the input automatically from the huge training data, which usually work better than hand-craft features.

Methods
Information Extraction and Image Generation
Convolutional Neural Network
The architecture of the proposed
The output
Experimental
Training
Results
Comparison

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.