ECE 110/120 Honors Lab Section : Infrared Camera Road following Car

Statement of Purpose

This project aims to create an autonomous vehicle that tracks and follows road surfaces and avoids obstacles. A big issue with using traditional cameras is that it is unable to work effectively in dim environments like at night. One solution to solve that issue would be to include an illumination source on the vehicle. Another solution would be to integrate other forms of sensors onto the vehicle. We will address this problem by using infrared thermal cameras. Infrared thermal cameras are light and color independent meaning daylight and color similarity between the path and its surroundings do not affect its performance. 

Goals:

• Implement thermal camera
• Detect and follow roads
• Detect and avoid obstacles

Background Research

After our initial brainstorming, we have found out that there can be various applications of infrared road-following car. For example, autonomous delivery robots can be the one option that we can apply to our project extensively. While most of the related products or projects are utilizing an infrared camera as an option to detect/avoid certain objects, we are looking forward to using the camera as an operation and driving system of the car that controls the car and possibly apply the system to the real-world application. We will be using temperature differences in the environment to identify the road, establish an ideal camera location to observe the difference, and make the car to follow specific regions. If time allows, we are planning to elaborate on our project for more practical usage. 

Block Diagram / Flow Chart

Block Diagram:

Flow Chart:

System Overview

The block diagram reflects the simple data and flow. The data in the Infrared Camera is sent to the MCU. The MCU does the image processing and determines which motor needs to be on and how long. This is done by sending a signal to the motor driver, which then sends power to the motors.

Despite the simple block diagram, the flow chart shows that most of the complexity resides in the MCU as it needs to process the received IR image and determine where the road is and where the vehicle is relative to the road. Every conditional statement represents a series of image processing operations that determine whether the vehicle is on the road or not.

Parts

Possible Challenges

A challenge we anticipate would be the math needed for the vehicle to determine where the path is and to filter out false positives. This means the microcontroller needs to be able to reject hotspots and determine whether what it is on is a road or not.

Another challenge might be the sensitivity and field of view of the infrared camera. The camera's sensitivity needs to be high enough to detect the difference in sidewalk temperature and surrounding temperatures. And the camera needs to be set up in a position where it is able to have a clear field of view of the sidewalk and its surroundings, infrared cameras are sensitive and require the proper focus range and distance to work properly

Another challenge that might arise is while using an infrared camera would be glass buildings. Depending on how we have the camera set up buildings with glass might cause problems as infrared cameras often face issues when glass is present as glass often reflects heat and can skew the image we are trying to obtain.

References

Davies, A., 2022. Heat-Seeking Cameras Could Help Keep Self-Driving Cars Safe. [online] Wired. Available at: <https://www.wired.com/story/self-driving-cars-thermal-image-cameras/> [Accessed 20 February 2022].

Scott King, P., 2022. Electrical Equipment and Challenges with Infrared Thermography. [online] Hallam-ics.com. Available at: <https://www.hallam-ics.com/blog/electrical-equipment-and-challenges-with-infrared-thermography> [Accessed 20 February 2022].