Event Detection and Imaging System for Vehicle Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Asynchronous imaging devices, like dynamic vision sensors, have lower resolution due to larger pixel sizes, making it difficult to accurately recognize objects despite detecting events, such as changes in luminance on a road surface, while mounted on a vehicle.
Innovation Solution
An imaging system comprising an event detection device that detects luminance changes exceeding a threshold, an imaging device capturing images at a fixed frame rate, and a control section that acquires image information of the event region, allowing for more accurate object recognition by switching from low-resolution event detection to high-resolution imaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If an event detection device uses a pixel configuration including an event detection section for each pixel, then it can detect luminance changes (events) effectively, but the pixel size becomes larger and resolution becomes lower
Solution Approach 1:
The imaging system is divided into two functional segments: an event detection device for detecting luminance changes and an imaging device for capturing high-resolution images. This segmentation allows each device to be optimized for its specific function, resolving the contradiction between event detection capability and object recognition accuracy
Solution Approach 2:
The control section acts as an intermediary that receives event detection results and triggers the imaging device to capture images of specific regions. This mediator coordinates between the low-resolution event detection and high-resolution imaging, enabling accurate object recognition only when needed
2Speed
If the event detection device captures images continuously at high frame rate, then it can monitor dynamic environments effectively, but the power consumption increases and processing load becomes heavier
Solution Approach 1:
Instead of continuous imaging, the system uses periodic action by triggering the imaging device only when events are detected. The event detection device operates continuously at high frame rate to monitor changes, while the imaging device activates periodically based on event triggers, reducing overall power consumption while maintaining monitoring effectiveness
Solution Approach 2:
The event detection device performs preliminary detection of luminance changes before triggering the main imaging device. This preliminary action filters out unnecessary imaging operations, allowing the system to maintain high monitoring response speed while significantly reducing power consumption by activating the high-power imaging device only when relevant events occur
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system enables accurate object recognition by leveraging the event detection device's ability to detect events and the imaging device's higher resolution, improving the detection of road surface conditions and objects, even in dynamic environments.
Implementation Method 1
a luminance change of a pixel that photoelectrically converts incident light exceeds a predetermined threshold
Data Source
AI summary
An imaging system of the present disclosure includes an event detection device that detects, as an event, that a luminance change of a pixel that photoelectrically converts incident light exceeds a predetermined threshold, an imaging device that captures an image at a fixed frame rate, and a control section that controls the event detection device and the imaging device, and is mounted on and used in a moving body. Then, the control section performs control to acquire image information of a region including the event by the imaging device in response to the event detection device detecting the event. Furthermore, an object recognition system of the present disclosure detects an event by the event detection device under the control of the control section, acquires image information of a region including the event by the imaging device, and performs object recognition based on the acquired image information.