3 Minutes
BYD files patent for machine vision system that spots animals under cars
BYD has filed a new patent for a machine vision system designed to detect living creatures lurking beneath parked vehicles. The technology shows how the Chinese automaker is expanding computer vision beyond driver assistance and battery management, bringing it into active vehicle safety and parking protection.
How the system works
At the core of BYD's proposal is an image-based reference map. When the vehicle is turned off, cameras capture a baseline image of the undercarriage. Subsequent live frames are compared against that stored reference to identify regions that have changed. Instead of continually scanning every pixel, the system first isolates areas that differ from the baseline and focuses compute resources there.
The patent specifies that fixed structures such as suspension components, battery housings, and aerodynamic panels are ignored to reduce false positives and cut unnecessary processing. Only newly introduced objects or movements under the vehicle trigger further analysis, where classification algorithms decide whether the change corresponds to an animal, a person, or debris.
Why this matters for safety and the environment
Detecting animals under parked cars is more than a convenience feature. It addresses a real-world safety concern for urban, suburban, and rural drivers alike. Small animals and even children can be hard to spot near low-clearance vehicles, and early detection can prevent injuries and vehicle damage.
The patent also acknowledges environmental and lighting challenges. Shadows, changing illumination, road grime, and uneven ground can all confuse traditional motion-detection systems. BYD's comparative approach — building a customized reference map per vehicle — aims to reduce those issues by filtering environmental noise before classification.
- Key benefits: targeted processing, fewer false alarms, adaptive to each vehicle
- Challenges addressed: shadows, lighting shifts, road dirt, surface variations
- Complementary tech: integrates with radar- or lidar-based cabin monitoring
Part of a wider sensor strategy
This patent complements a previous BYD invention for detecting forgotten occupants inside the cabin using radar data. Together, the two systems would monitor both the exterior and interior of a vehicle, forming a broader smart-sensor ecosystem that combines machine vision, radar processing, and advanced monitoring algorithms.
Quote highlight: 'By building a per-car reference map and focusing compute on changes, BYD seeks to make undercarriage detection practical and reliable in real conditions.'
This patent filing does not guarantee immediate deployment. There is no official timeline or confirmation that the technology will appear in production models. Still, the move signals BYD's continued investment in sensor fusion and automotive safety — an area likely to grow as automakers pursue smarter, safer vehicles worldwide.
Leave a Comment