Improving Keypoint Matching Using a Landmark-Based Image Representation

Motivated by the need to improve the performance of visual loop closure verification via multi-view geometry (MVG) under significant illumination and viewpoint changes, we propose a keypoint matching method that uses landmarks as an intermediate image representation in order to leverage the power of deep learning. In environments with various changes, the traditional verification method via MVG may encounter difficulty because of their inability to generate a sufficient number of correctly matched keypoints. Our method exploits the excellent invariance properties of convolutional neural network (ConvNet) features, which have shown outstanding performance for matching landmarks between images. By generating and matching landmarks first in the images and then matching the keypoints within the matched landmark pairs, we can significantly improve the quality of matched keypoints in terms of precision and recall measures. The proposed method is validated on challenging datasets that involve significant illumination and viewpoint changes, to establish its superior performance to the standard keypoint matching method.

Authors

• Xinghong Huang
• Zhuang Dai
• Weinan Chen
• Li He 0002
• Hong Zhang 0013

Keywords

• Lighting
• Proposals
• Visualization
• Standards
• Feature extraction
• Simultaneous localization and mapping
• Image representation
• Keypoint Matching
• Standard Method
• Precision And Recall
• Illumination Changes
• Intermediate Representation
• Representation In Order
• Loop Closure
• Viewpoint Changes
• Visual Verification
• False Positive
• Environmental Changes
• True Positive
• Rest Of The Paper
• Loss Of Generality
• Descriptive Characteristics
• Object Detection
• Bounding Box
• Image Pairs
• Changes In Appearance
• AlexNet
• Nearest Neighbor Search
• False Image
• True Pairs
• Matching Quality
• Pruning Techniques
• True Image
• Robot Navigation