A

ANTIPINCH TECHNOLOGY USING MACHINE LEARNING

Automotive anti-pinch is a safety system used in the power windows of a car. The technology prevents the winding up of the window if it senses any obstacle in the path of the glass.

A

ANTIPINCH TECHNOLOGY USING MACHINE LEARNING

Automotive anti-pinch is a safety system used in the power windows of a car. The technology prevents the winding up of the window if it senses any obstacle in the path of the glass.

The problem ANTIPINCH TECHNOLOGY USING MACHINE LEARNING solves

Anti-pinch technology is a safety system that is present in power windows
of modern vehicles. The power window is the window that uses an electric
motor to operate. As the name anti-pinch suggests, this window prevents the
accidental winding up of the power window. If the system senses any obstacle
in the path of the glass, it prevents the window glass from moving further up.
Thus, it prevents the possible injuries to the occupants.
This project deals with the anti-pinch technology using image processing.
The camera module is used to identify the presence of objects near the window.
The camera module continuously monitors the path of the glass window and if
there is any presence of obstacles the camera module sends the signal to relay
and thus preventing the power window from winding up.
The components used in this project is a NI mydaq and a camera module
integrated with it. This system turns on only when the power window is winding
up. LabVIEW is used to control the NI mydaq and python program is used to
detect the hand movement near the window
In case of window equipped with anti-pinch technology, the electric
motor is fitted with a relay and when the camera module senses any objects in
the path of the glass the module sends a signal to the relay and the relay starts
operating in reverse direction. As a result, the window comes down, avoiding
any injury to the occupant.

Challenges we ran into

Low light conditions and the proper alignment of the camera are the challenges we ran into

Discussion