Proyectos

On-chip Multispectral Image Processing and its Application to Advanced Driving Assitance Systems (ADAS)

Personal investigador:
Koldo Basterretxea (PI)
Periodo:
desde 2019 hasta 2021
Entidad financiadora:
Basque Goverment
Importe total:
39,066 €
Descripción:

The use of hyperspectral sensors can help to improve

scene understanding and object tracking system robustness,

since they provide richer information about materials than

conventional cameras. This extra information can be used

to better separate objects and backgrounds, improve object

tracking robustness and enhance the performance of image

segmentation algorithms. HSI is already being widely applied in some areas such as remote sensing in geoscience and,

more recently, in precision agriculture, medical imaging and

others. However, the use of HSI in application areas

that require autonomy of operation and easy integration in

moving platforms has been limited by traditional scanning

sensor setup requirements and also by restrictions in the

processing power of the accompanying computing hardware.

To-day, with the advent of new small-size snapshot cameras

that can provide hyperspectral images at the video rates,

it is possible to setup a HSI system in almost any moving

platform, and particularly in ground vehicles. Consequently,

how to efficiently process hyperspectral information onboard

a moving platform is currently an active research topic.