In nature, snakes can gracefully traverse a wide range of different and complex environments. Snake robots that can mimic this behaviour could be fitted with sensors and transport tools to hazardous or confined areas that other robots and humans are unable to access. In order to carry out such tasks, snake robots must have a high degree of awareness of their surroundings (i.e., perception-driven locomotion) and be capable of efficient obstacle exploitation (i.e., obstacle-aided locomotion) to gain propulsion. These aspects are pivotal in order to realise the large variety of possible snake robot applications in real-life operations such as fire-fighting, industrial inspection, search-and-rescue, and more. In this paper, we survey and discuss the state of the art, challenges, and possibilities of perception-driven obstacle-aided locomotion for snake robots. To this end, different levels of autonomy are identified for snake robots and categorised into environmental complexity, mission complexity, and external system independence. From this perspective, we present a step-wise approach on how to increment snake robot abilities within guidance, navigation, and control in order to target the different levels of autonomy. Pertinent to snake robots, we focus on current strategies for snake robot locomotion in the presence of obstacles. Moreover, we put obstacle-aided locomotion into the context of perception and mapping. Finally, we present an overview of relevant key technologies and methods within environment perception, mapping, and representation that constitute important aspects of perception-driven obstacle-aided locomotion.

No abstract

Deep learning methods have been successfully applied to image processing, mainly using 2D vision sensors. Recently, the rise of depth cameras and other similar 3D sensors has opened the field for new perception techniques. Nevertheless, 3D convolutional neural networks perform slightly worse than other 3D deep learning methods, and even worse than their 2D version. In this paper, we propose to improve 3D deep learning results by transferring the pretrained weights learned in 2D networks to their corresponding 3D version. Using an industrial object recognition context, we have analyzed different combinations of 3D convolutional networks (VGG16, ResNet, Inception ResNet, and EfficientNet), comparing the recognition accuracy. The highest accuracy is obtained with EfficientNetB0 using extrusion with an accuracy of 0.9217, which gives comparable results to state-of-the art methods. We also observed that the transfer approach enabled to improve the accuracy of the Inception ResNet 3D version up to 18% with respect to the score of the 3D approach alone.

This article deals with the 2D image-based recognition of industrial parts. Methods based on histograms are well known and widely used, but it is hard to find the best combination of histograms, most distinctive for instance, for each situation and without a high user expertise. We proposed a descriptor subset selection technique that automatically selects the most appropriate descriptor combination, and that outperforms approach involving single descriptors. We have considered both backward and forward mechanisms. Furthermore, to recognize the industrial parts a supervised classification is used with the global descriptors as predictors. Several class approaches are compared. Given our application, the best results are obtained with the Support Vector Machine with a combination of descriptors increasing the F1 by 0.031 with respect to the best descriptor alone.

Detection and description of keypoints from an image is a well-studied problem in Computer Vision. Some methods like SIFT, SURF or ORB are computationally really efficient. This paper proposes a solution for a particular case study on object recognition of industrial parts based on hierarchical classification. Reducing the number of instances leads to better performance, indeed, that is what the use of the hierarchical classification is looking for. We demonstrate that this method performs better than using just one method like ORB, SIFT or FREAK, despite being fairly slower.