LiDAR is a measurement signal method that measures distance to a target by illuminating that target with a pulsed laser light and measuring the reflected pulses with a sensor. Differences in laser return times and wavelength can then be used to make digital 3D representations of the target.
There has been considerable work on object classification for autonomous vehicle navigation. Most of the works use camera vision to capture 2D images of vehicles and other vulnerable road users such as pedestrians and bicy- clists. However, camera vision alone may not be able to provide important depth information to detect and track objects with the level of reliability needed for safe driving. LiDAR enables generation of 3D images from a single sweep. Multiple sweeps can be used to generate information about velocities and distances.
In our company YOTASYS, we train the signal data using Neural Networks after any required reprocessing or feature extraction. Neural Networks use the processing of the brain as a basis to develop algorithms that can be used to model complex patterns or prediction problems. Neural networks have the ability to learn from initial inputs and their relationships, generalize the model and predict on unseen data.
Our goal is to create a system that can efficiently and accurately predict and classify ahead of time the objects in its surroundings, so as to aid the self-automated robots or systems in maneuvering through safely.
Please refer to our products page for more details on realised applications.