Extreme Low Light Image Sensor Performance - On the Road

by  ON Semiconductor  - 04-02-2019 

At ON Semiconductor we know that imaging applications need to perform in challenging conditions and must operate 100% of the time. So we put our image sensors through tough, real-world tests so you can have confidence in the products you get from us.

Watch as ON Semiconductor engineers ensure you're safety when testing automotive imaging reliability that is critical for autonomous driving in extreme low light conditions.

Full Transcript

It is difficult to test the extremely low light sensitivity performance of automotive image sensors, especially in the realm of less than 0.5 lux. This equates to a real life scenario like a dark road with no street lights and no taillights. This type of lighting situation is an Autonomous driving nightmare, as the scene dynamically changes as a vehicle moves past different objects. Noise characteristics of the images change constantly and impacts how objects are viewed and interpreted by the image sensor.

Autonomous vehicles have image sensors located all around the vehicle, not all of them have the benefit of having a light source near them. Even at a very low light situation, we still have issues with motion artifacts. Another concern can be how the image sensor does or does not interpret the data it receives. For example, it is important for an image sensor to be able to determine via headlight separation, whether the vehicle is a car, a truck, a bike, a scooter, etc.

We anticipate real-life situations, and then put our image sensors to the test, whether they be challenging lighting situations, adverse weather conditions, or broad temperature ranges.

We test our sensors in the same scenarios and situations that our customers do to ensure they work, not only in the lab, but in real life too.

We have done the testing for you. For more information on our Automotive Image Sensors, please click here


Tags:Automotive, CMOS Image Sensor
What's your reaction?
0 0 0 0 0

Post a Comment

Your email address will not be published. Required fields are marked *




Rules for Commenting


Zoran Najdovski 2021-11-25 00:00 Comment rating: Vote down 0 Vote up
Hi ONSEMI team, We are currently scoping a new research project at Deakin University in Australia for a driver warning system based on object detection through machine learning. We will be collecting data from 4 cameras on the vehicle (viewing range approx. 300 meters) in the first phase using a laptop PC. In subsequent phases we will be deploying this on an NVIDIA Jetson AGX and reducing the number of cameras to 1. The system is primarily targeting low-light conditions. Could you please recommend the best camera sensor for our application? As you would appreciate, a global shutter solution is more desirable. For a multi-camera data capture setup we would prefer a solution (even third party using your recommended sensor) which supports GMSL interfacing. Kind Regards, Zoran Dr. Zoran Najdovski Associate Professor, Deakin-DSI Liaison Manager Institute for Intelligent Systems Research and Innovation (IISRI) Deakin University, 75 Pigdons Road, Waurn Ponds, Victoria, 3216, Australia Tel: +61 3 5227 2430 Fax: +61 3 5227 1046 Email: zoran.najdovski@deakin.edu.au http://www.deakin.edu.au/research/iisri
Your request has been submitted for approval.
Please allow 2-5 business days for a response.
You will receive an email when your request is approved.
Request for this document already exists and is waiting for approval.