An Improved Deep Residual Network-Based Semantic Simultaneous Localization and Mapping Method for Monocular Vision Robot

Figure 3

The flowchart of the improved key frame insertion mechanism.