1. Robot vision sensors
Emerging in the late 1950s, machine vision has developed rapidly and is one of the most important sensors in robotics. Starting in the 1960s, machine vision initially processed the world of building blocks, later expanding to handle the real world outdoors. Practical vision systems emerged after the 1970s. Vision generally includes three processes: image acquisition, image processing, and image understanding. Relatively speaking, image understanding technology is still quite underdeveloped.
2. Force sensor
In terms of installation location, robot force sensors can be divided into joint force sensors, wrist force sensors, and finger force sensors. International research on wrist force sensors began in the 1970s, with major research institutions including DRAPER Laboratories, SRI Research, and IBM in the United States, and Hitachi and the University of Tokyo in Japan.
3. Touch sensor
As a complement to vision, touch can perceive the surface properties and physical characteristics of a target object: softness, hardness, elasticity, roughness, and thermal conductivity, etc. Tactile research began in the early 1980s and yielded significant results by the early 1990s.
4. Proximity sensor
The purpose of this research is to enable robots to know the proximity of targets (obstacles) during movement or operation. Mobile robots can avoid obstacles, and manipulating robots can avoid the impact of the gripper on the target due to excessive approach speed.
5. Hearing sensor
(1) Voice recognition system for specific individuals
Speaker-specific speech recognition methods store the feature matrix of each syllable in the voice of a pre-designated person, forming a standard template, and then perform matching. It first memorizes one or several speech features, and the content of the designated person's speech must be a limited set of pre-defined sentences. A speaker-specific speech recognition system can then identify whether the speaker is the pre-designated person and which sentence they are speaking.
(2) Person-independent speech recognition system
Speaker-independent speech recognition systems can be broadly categorized into language recognition systems, word recognition systems, and digit sound (0-9) recognition systems. Speaker-independent speech recognition methods require training on a representative group of speakers to identify commonalities in the sounds of the same words. This training is often open-ended, allowing for continuous system refinement. During system operation, the feature matrices of received sound signals are calculated using the same method and then compared with standard patterns. The system identifies which template it matches or is similar to, thereby determining the meaning of the signal.
