Online Multispectral Neuron Tracing

In this paper, we propose an online neuron tracing method for multispectral images with unique modules, where no offline training is required. Our method is online trained to update our enhanced discriminative correlation filter to conglutinate the tracing process. This offline-training-free schema differentiates us from other training-dependent approaches like deep learning methods. Besides, compared to other tracing methods requiring complicated setup such as for clustering and graph multi-cut, our approach is much easier to be applied to new images. In fact, it only needs a starting bounding box of the tracing neuron, significantly reducing users' configuration effort. Our extensive experiments show that our training-free and easy-configured methodology allows fast and accurate neuron reconstructions in multispectral images. Code is available at https://github.com/tuffr5/online-tracing.