Abstract
Simultaneous Localization and Mapping(SLAM) is used to solve the problem of mobile robot navigation and map building in an unknown environment. Loop closure detection is the key part of SLAM, which determines the precision and stability of SLAM to a great extent. Closed loop detection is affected by changes in light and dynamic environment. In order to improve the accuracy of the loop closure detection, this paper introduces the model of deep learning to transform the visual SLAM process. On the basis of Sparse Autoencoder(SAE), the activation function of the output layer is changed to the identity function to solve the problem that the input sample needs to be scaled and does not apply to the color image. Then, the input image and the trained network do convolution and pooling operation. Not only can significantly reduce the feature dimension, but also can improve the effect of feature description. The test conducted by using the open data set shows that the proposed method can effectively solve the problem of visual SLAM loop closure detection.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
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.