Event-Based Camera Height Measurement via Edge Speed Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring the height of objects in a moving stream, such as those on a conveyor belt, are either expensive, complex, or provide height values too late for applications like focus tracking, and event-based cameras lack the capability to measure depth values.
Innovation Solution
An event-based camera with a matrix of pixel elements that detect changes in intensity, allowing for the determination of object height by analyzing the speed of object edges as they move across the sensor, which is converted into height using the speed of the object image on the sensor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If an additional sensor such as a laser scanner is associated with the camera to measure geometry, then height measurement capability is improved, but hardware effort and installation effort increase
Solution Approach 1:
The patent makes the existing image sensor perform multiple functions: it not only captures image data for object identification but also measures height information by detecting the temporal progression of object edges across adjacent pixels. This eliminates the need for additional dedicated sensors like laser scanners, resolving the contradiction between measurement capability and device complexity.
2Measurement precision
If 3D cameras are used to measure distances and heights, then height measurement capability is improved, but cost increases and lateral resolution quality deteriorates
Solution Approach 1:
The patent enables the conventional 2D image sensor to self-determine height information by utilizing its existing pixel array and temporal sampling capability. The sensor processes its own output data to extract depth information through the motion of object edges across adjacent pixels, eliminating the need for more expensive 3D cameras while maintaining or improving lateral resolution.
3Measurement precision
If height is acquired from a plurality of image records in a time interval using stereo camera principle, then height measurement capability is improved, but system complexity increases and height value is available too late for focus tracking
Solution Approach 1:
The patent performs height measurement in advance during the object's passage through the field of view by continuously monitoring adjacent pixels. The height information is extracted from the temporal pattern of edge detection across multiple pixels before the object leaves the field of view, making the measurement available immediately for focus tracking without the delays associated with post-processing multiple images.
4Ease of operation
If conventional cameras are used to capture images at regular refresh rate, then image acquisition is simplified, but depth value and height measurement capability are lost
Solution Approach 1:
The patent introduces dynamic temporal sampling at the pixel level, where each pixel independently detects and reports changes in intensity at different times. This dynamic approach allows the system to capture motion information and calculate depth from the temporal patterns of edge detection across adjacent pixels, adding height measurement capability while maintaining the simplicity of conventional image acquisition.
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
Enables simple and robust height measurement with reduced development effort and calculation capacity, providing high temporal resolution and eliminating motion blur, while being insensitive to extraneous light and reducing data processing requirements.
Implementation Method 1
each pixel individually checks whether it determines a change in intensity
Implementation Method 2
the speed of the object image on the sensor, which is converted into height
Data Source
AI summary
A camera is provided that has an image sensor having a plurality of pixel elements for recording image data of an object stream moved in a direction of movement relative to the camera and having objects of a height, and a control and evaluation unit that is configured to determine the height of a respective object with reference to the image data. The image sensor here is an event-based image sensor and the control and evaluation unit is configured to detect a measurement series of points in time at which pixel elements adjacent in the direction of movement consecutively register an event to determine a speed of an object image of the object on the image sensor from said measurement series and from this the height.