Abstract
Light detection and ranging (LiDAR) is a frequently used technique of data acquisition and it is widely used in diverse practical applications. In recent years, deep convolutional neural networks (CNNs) have shown their effectiveness for LiDAR-derived rasterized digital surface models (LiDAR-DSM) data classification. However, many excellent CNNs have too many parameters due to depth and complexity. Meanwhile, traditional CNNs have spatial redundancy because different convolution kernels scan and store information independently. SqueezeNet replaces a part of 3 × 3 convolution kernels in CNNs with 1 × 1 convolution kernels, decomposes the original one convolution layer into two layers, and encapsulates them into a Fire module. This structure can reduce the parameters of network. Octave Convolution (OctConv) pools some feature maps firstly and stores them separately from the feature maps with the original size. It can reduce spatial redundancy by sharing information between the two groups. In this article, in order to improve the accuracy and efficiency of the network simultaneously, Fire modules of SqueezeNet are used to replace the traditional convolution layers in OctConv to form a new dual neural architecture: OctSqueezeNet. Our experiments, conducted using two well-known LiDAR datasets and several classical state-of-the-art classification methods, revealed that our proposed classification approach based on OctSqueezeNet is able to provide competitive advantages in terms of both classification accuracy and computational amount.
Highlights
Light detection and ranging (LiDAR) technology is an active remote sensing measurement technology which can obtain ground object information by emitting laser to the target [1]
Ghamisi et al proposed a joint classification method which is based on extinction profiles (EPs) features and convolutional neural network (CNN) to improve the classification accuracy [14]
We propose a novel dual neural architecture, OctSqueezeNet, which combines SqueezeNet with Octave Convolution (OctConv), which improves LiDAR-DSM data classification accuracy with less structural parameter memory
Summary
Light detection and ranging (LiDAR) technology is an active remote sensing measurement technology which can obtain ground object information by emitting laser to the target [1]. Ghamisi et al proposed a joint classification method which is based on extinction profiles (EPs) features and convolutional neural network (CNN) to improve the classification accuracy [14]. Wang et al combined morphological profiles (MPs) and CNN to provide more features for LiDAR-DSM classification [16]. The features of adjacent pixels in one image have similarities; the convolution kernel in traditional CNNs sweeps each location and stores its own feature description independently It ignores spatial consistency, so that the feature maps have a large amount of redundancy in the spatial dimension. We propose a novel dual neural architecture, OctSqueezeNet, which combines SqueezeNet with OctConv, which improves LiDAR-DSM data classification accuracy with less structural parameter memory. We replaced the original convolutional layer in OctConv with the Fire modules in SqueezeNet to compensate for the shortcomings of SqueezeNet, which had less extracted feature information due to lots of 1 × 1 convolution kernel
Sign-in/Register to view highlights & summary 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 callDisclaimer: 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.
