Multispectral background subtraction with deep learning

In this paper, we follow the trend of deep learning and make an attempt to investigate the potential benefit of using multispectral images via convolutional neural networks for background subtraction task. The major contributions of this work lie in two aspects, based on the impressive algorithm FgSegNet_v2. Firstly, we extract three channels out of the seven of the FluxData FD-1665 multispectral dataset to match the number of input channels of the VGG16 deep model. Some combinations of three-channel based multispectral images perform better than RGB images. Secondly, a new convolutional encoder is designed to use all the multispectral channels available to further explore the information of multispectral images. The results outperform the RGB images and also other approaches using the same multispectral dataset.

Automated summary

This paper investigates the potential benefits of using multispectral images for background subtraction task using convolutional neural networks, with impressive results outperforming RGB images and other approaches in the experiments. A new convolutional encoder is designed to explore the information of multispectral images, demonstrating the advantages of using multispectral data with deep learning techniques.