| Robust Building Footprint Extraction from Big Multi-Sensor Data using Deep Competition Network |
| Paper ID : 1094-SMPR |
| Authors: |
|
Mehdi Khoshboresh Masouleh *, Mohammad Reza Saradjian School of Surveying and Geospatial Engineering, College of Engineering, University of Tehran, Tehran, Iran |
| Abstract: |
| Building footprint extraction (BFE) from multi-sensor data such as optical images and 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. The proposed DCN in this study is based on modification of our previous model on BFE from multi-sensor data. DCN is a deep 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 (feature) learning. DCN is trained and tested in a big multi-sensor dataset obtained from the state of Indiana in 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. |
| Keywords: |
| Building Extraction, Big Multi-Sensor Data, Deep Learning, Convolutional Network, Cloud Computing, Colab |
| Status : Conditional Accept (Poster) |
