Scene matching areas classification based on PCANet and MLP

Kai Sun; Liang Pan; Weilin Yuan

doi:10.1117/12.2544155

27 November 2019 Scene matching areas classification based on PCANet and MLP

Kai Sun, Liang Pan, Weilin Yuan

Proceedings Volume 11321, 2019 International Conference on Image and Video Processing, and Artificial Intelligence; 1132125 (2019) https://doi.org/10.1117/12.2544155
Event: The Second International Conference on Image, Video Processing and Artifical Intelligence, 2019, Shanghai, China

Abstract

Scene matching aided navigation is mainly used in autonomous navigation of aircraft. In scene matching field, scene matching areas selecting is a great challenge. The traditional methods focus on extracting image features and building a model to fit the relationship between image features and matching suitability indicators. We propose a new method combing principal component analysis network (PCANet) and multi-layer perception (MLP) to select scene matching areas for the first time. Firstly, we built a dataset based on images captured by TerraSAR-X satellite. Secondly, we extract information of each image by PCANet and generate label based on matching probability. Finally, MLP is used to automatically fit the mapping relation between image and matching suitability. The proposed method avoids the steps of extracting features manually and improves the performance in different task. The method proposed in this paper performs better than convolutional neural network (CNN).

Citation Download Citation

Kai Sun, Liang Pan, and Weilin Yuan "Scene matching areas classification based on PCANet and MLP", Proc. SPIE 11321, 2019 International Conference on Image and Video Processing, and Artificial Intelligence, 1132125 (27 November 2019); https://doi.org/10.1117/12.2544155

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available