Machine-Learning-Based Visual Objects’ Distances Evaluation: a Comparison of ANFIS, MLP, SVR and Bilinear Interpolation Models
Résumé
Spatial characterization of objects is a key-step for robots’ awareness about the surrounding environment in which they are supposed to evolve and for their autonomy within that environment. Within this context, this chapter deals with visual evaluation of objects’ distances using Soft-Computing based approaches and pseudo-3D standard low-cost sensor, namely the Kinect. However, although presenting appealing advantages for indoor environment’s perception, the Kinect has not been designed for metrological aims. The investigated approach offers the possibility to use this low-cost pseudo-3D sensor for the aforementioned purpose by avoiding 3D feature extraction and thus exploiting the simplicity of the only 2D image’ processing. Experimental results show the viability of the proposed approach and provide comparison between different Machine-Learning techniques as Adaptive-network-based fuzzy inference (ANFIS), Multi-layer Perceptron (MLP), Support vector regression (SVR) and Bilinear Interpolation (BLI).