eRTIS: A Fully Embedded Real Time 3D Imaging Sonar Sensor for Robotic Applications

Abstract

Many popular advanced sonar systems provide accurate and reliable measurements containing crucial info needed by robotic applications such as range, bearing and reflection strength of the objects in the field of view. While these sensor systems provide these crucial pieces of information accurately, they are often limited by a lack of processing power and/or size which leads to them needing an external computing device to process all the information generated by the microphone array on the sensor. In this paper we present two versions of a novel fully embedded 3D sonar sensor which have different sensing architectures which enable 3D perception for robotic application in harsh conditions using ultrasound at low cost. Experimental results taken from an office environment will show the 3D localization capabilities and performance of the sensor, showing the sensor has a large field-of-view (FoV) with accurate 3D localization combined with real-time capabilities.

Predicting LiDAR Data From Sonar Images

“…It is worth emphasizing that the model is trained using real-world measurements, instead of using a simulated approach. In contrast to the sonar sensor used in this work, we used a novel sonar sensor created within our research group which provides a wide field-of-view with accurate 3D localization using 32 microphones [24], [12].…”

“…In the discussion above we argued using a deep neural network as an appropriate approach for the prediction of LiDAR point clouds based on the measurements of a sonar sensor. We interpreted the problem as an inverse imagining problem where the goal is to reconstruct LiDAR data using the measurements of the eRTIS sonar sensor developed by our research group [12]. We use the sonar data at the input of our network, the measurements are processed into energyscape images which similarly represent the environment as the LiDAR data.…”

“…Only edges and the echo at the normal of the surface and the sensor are measured. The energyscapes are obtained by a beamforming technique on the echoes measured by the eRTIS sensor, for more information about the sonar sensor and the processing to obtain the energyscape images the reader is redirected to [12] or [24].…”

“…The measurements produced by these sensors in non-optimal sensing environment should not be solely trusted for making navigation decisions. In previous work, our research group (CoSys-Lab) has developed a novel 3D in-air sonar sensor (eRTIS) [12], which is capable of generating accurate 3D images of the environment using ultrasound. In this work, we propose a method for learning a transformation of sonar measurements into LiDAR point clouds.…”

“…The goal of multimodal sensor fusion is to use a set of sensors that complement each other in different situations, one modality can perform worse than another, but the overall performance should be less affected. We argue that the eRTIS in-air sonar sensor [12] is an excellent addition to multimodal sensing systems that should be capable of accurately sensing in difficult and harsh environments. In bad visibility situations, the LiDAR and camera sensors will perform worse or fail, the eRTIS sensor is still able to perform in this situation.…”