Semantic Segmentation to Develop an Indoor Navigation System for an Autonomous Mobile Robot
| Autor: | Ekaitz Zulueta Guerrero, Daniel Teso Fernández de Betoño, Aitor Sáenz Aguirre, Unai Fernández Gámiz, Ander Sánchez Chica |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Computer science
General Mathematics 0211 other engineering and technologies 02 engineering and technology Unet ResNet obstacle detection Regional development Human–computer interaction Agency (sociology) 0202 electrical engineering electronic engineering information engineering Computer Science (miscellaneous) Production (economics) fully convolutional networks Segmentation Engineering (miscellaneous) 021101 geological & geomatics engineering lcsh:Mathematics Navigation system indoor navigation Segnet Mobile robot Economic support lcsh:QA1-939 semantic segmentation autonomous mobile robot Robot 020201 artificial intelligence & image processing |
| Zdroj: | Addi. Archivo Digital para la Docencia y la Investigación Universidad de Cantabria (UC) Mathematics Volume 8 Issue 5 Mathematics, Vol 8, Iss 855, p 855 (2020) instname Addi: Archivo Digital para la Docencia y la Investigación Universidad del País Vasco |
| Popis: | In this study, a semantic segmentation network is presented to develop an indoor navigation system for a mobile robot. Semantic segmentation can be applied by adopting different techniques, such as a convolutional neural network (CNN). However, in the present work, a residual neural network is implemented by engaging in ResNet-18 transfer learning to distinguish between the floor, which is the navigation free space, and the walls, which are the obstacles. After the learning process, the semantic segmentation floor mask is used to implement indoor navigation and motion calculations for the autonomous mobile robot. This motion calculations are based on how much the estimated path differs from the center vertical line. The highest point is used to move the motors toward that direction. In this way, the robot can move in a real scenario by avoiding different obstacles. Finally, the results are collected by analyzing the motor duty cycle and the neural network execution time to review the robot&rsquo s performance. Moreover, a different net comparison is made to determine other architectures&rsquo reaction times and accuracy values. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|