Object Detection and Recognition for Assistive Robots: Experimentation and Implementation

Abstract

Abstract-Technological advances are currently being directed to assist the human population in performing ordinary tasks in everyday settings. In this context, a key issue is the interaction with objects of varying size, shape and degree of mobility. Consequently, autonomous assistive robots must be provided with the ability to process visual data in real time so that they can react adequately for quickly adapting to changes in the environment. Reliable object detection and recognition is usually a necessary early step to achieve this goal. In spite of significant research achievements, this issue still remains a challenge when real-life scenarios are considered. In this paper, we present a vision system for assistive robots that is able to detect and recognise objects from a visual input in ordinary environments in real time. The system computes colour, motion and shape cues combining them in a probabilistic manner to accurately achieve object detection and recognition, taking some inspiration from vision science. In addition, with the purpose of processing the input visual data in real-time, a Graphical Processing Unit (GPU) has been employed. The presented approach has been implemented and evaluated on a humanoid robot torso located at realistic scenarios. For further experimental validation, a public image repository for object recognition has been used, allowing a quantitative comparison with respect to other state-of-the-art techniques when real-world scenes are considered. Finally, a temporal analysis of the performance is provided with respect to image resolution and number of target objects in the scene.

Simplificación de la tarea de picking industrial para el diseño de un sistema robótico

“…En general, estos desafíos giran en torno a dos grandes problemas relacionados con la visión y el agarre. El primero de ellos (el reconocimiento de objetos) es un problema clásico [10], [15]. En el segundo problema están implicadas varias disciplinas dentro del campo de la robótica: manipulación complaciente [13], control cinemático de robots bimanipuladores [12], planificación de trayectorias [11], [17] y técnicas de agarre [2].…”

Generic Object Discrimination for Mobile Assistive Robots Using Projective Light Diffusion

“…As a result, processing the entire volume of sensory data may hinder a timely robot response to certain object classes, therefore exploiting both color and geometry is generally more computationally restricting. The issue had been early recognized, nevertheless, more recent works [5] suggest that real-time as well as reliable object perception remains challenging.…”

IoT Driven Ambient Intelligence Architecture for Indoor Intelligent Mobility

“…Personal assistive robots that help the elderly to be more independent [3], or robots that help the blind [4] or robots that assist with rehabilitation [5] are some of the emerging applications of personal assistive robots. The robots will assist humans in many day to day tasks such as personal hygiene, mobility guidance [6], dressing support [7], feeding support [8], and rehabilitation support. The opportunities with assistive robots are numerous and it is expected that personal robots will form an integral part of future smart living environments.…”