General-Purpose Deep Learning Detection and Segmentation Models for Images from a Lidar-Based Camera Sensor

摘要

Over the last decade, robotic perception algorithms have significantly benefited from the rapid advances in deep learning (DL). Indeed, a significant amount of the autonomy stack of different commercial and research platforms relies on DL for situational awareness, especially vision sensors. This work explored the potential of general-purpose DL perception algorithms, specifically detection and segmentation neural networks, for processing image-like outputs of advanced lidar sensors. Rather than processing the three-dimensional point cloud data, this is, to the best of our knowledge, the first work to focus on low-resolution images with a 360° field of view obtained with lidar sensors by encoding either depth, reflectivity, or near-infrared light in the image pixels. We showed that with adequate preprocessing, general-purpose DL models can process these images, opening the door to their usage in environmental conditions where vision sensors present inherent limitations. We provided both a qualitative and quantitative analysis of the performance of a variety of neural network architectures. We believe that using DL models built for visual cameras offers significant advantages due to their much wider availability and maturity compared to point cloud-based perception.