Sensor Input Learning for Time-of- Flight Scan Laser

Abstract

The paper presents the most important stages of product development for a new 2D TOF laser
range sensor to be implemented in an autonomous platform. Some essential Scan-Laser characteristics
are: measure in real-time, approximately, medium distances, works in high frequency, does not weigh
much and is available at affordable prices. It has good power consumption and covers large scanning area
with resolution lower than 1cm, all these makes if fit for various applications. Measurements of the
distances to environmental objects is based on the principle Time-of- Flight (TOF). It consists in
determining the distances by time span of transmitted and, consequently, received light waves. Mitigation
of noise and disturbances effects is done by further processing collected data in learned neural networks
model and statistical inference. Important characteristics of laser sensor prototype, its working principle
for distance measurements, as well as some application and experimental results do really point out the
potential and benefits of this prototype.