Digital Vision Sensor with Logarithmic Time Encoding for High Dynamic Range
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vision sensors face limitations in dynamic range, leading to low sensitivity during varying light conditions, and existing solutions either suffer from analog noise or reduced resolution.
Innovation Solution
A digital vision sensor with a logarithmic time measurement system, where each pixel generates a voltage proportional to illumination, measures the time to reach a common reference voltage, and encodes this time with a binary code for invariant contrast computation, independent of light intensity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a conventional vision sensor with linear integration is used, then the sensor can operate over a large dynamic range of light illumination, but the sensitivity is low during low illumination and the dynamic range is limited
Solution Approach 1:
The patent implements a dual integration path system where the integration time is dynamically adjusted based on the detected light intensity. A first integration path with a first integration time is used for one dynamic range portion, while a second integration path with a second integration time is used for another dynamic range portion. This dynamic switching allows the sensor to optimize sensitivity for low illumination while maintaining the ability to handle high illumination levels, thereby resolving the contradiction between dynamic range and sensitivity.
2Illumination intensity
If a logarithmic value of the output quantity is provided to improve dynamic range, then the dynamic range is improved, but the sensor becomes particularly sensitive to noise and complex to implement
Solution Approach 1:
The patent replaces the analog logarithmic computation approach with a digital implementation. Instead of using analog components to compute logarithmic values (which are sensitive to noise), the system uses digital processing of the integration results. The microprocessor calculates the logarithm of the ratio between the two integration values, which provides the extended dynamic range while being much less sensitive to noise and easier to implement with standard digital components.
3Measurement precision
If local computation of contrast is performed by means of an analog multiplier, then the computation is carried out in an analog way with accuracy independent of light illumination level, but the resolution of the sensor is lower and analog noise issues persist
Solution Approach 1:
The patent replaces the analog multiplier-based contrast computation with a digital implementation. The system stores the integration results in digital memory and performs contrast calculations using digital processing. This substitution maintains the accuracy of contrast computation independent of illumination level while significantly improving sensor resolution and eliminating analog noise issues associated with the multiplier approach.
4Measurement precision
If the integration time is extended to improve sensitivity during low illumination, then the sensitivity is improved, but the measurement time increases and the dynamic range is reduced
Solution Approach 1:
The patent segments the integration process into two parallel paths with different integration times. The first integration path uses a longer integration time optimized for low illumination conditions, while the second integration path uses a shorter integration time for high illumination conditions. By segmenting the integration process and selectively using the appropriate path based on light levels, the system improves sensitivity during low illumination without permanently increasing the measurement time for all conditions.
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 solution provides a high dynamic range imaging capability with improved robustness and accuracy, allowing for contrast computation that is invariant to light illumination levels, reducing the need for storage and computation while maintaining sensitivity.
Implementation Method 1
each pixel having a photodetector and intended to receive a light beam representing a visual scene
Implementation Method 2
measuring the time which the voltage resulting from the integration of a photo-current on a capacitance takes for reaching a threshold
Data Source
AI summary
A digital vision sensor includes a pixel array (10) for receiving a light beam representing a visual scene. The sensor includes: a reference voltage Vref generator; in each pixel, an element for generating, during an integration phase, a voltage Vp that is proportional to the received lighting and a comparator (30) for determining the moment when the voltage Vp thereof reaches the reference voltage; a logarithmic period generator for generating a clock having a period that increases proportionally to the time elapsed from the beginning of the integration; a device for counting the number of clock periods elapsed from the beginning of the integration and providing a result in the form of a binary code; a static RAM memory word (50) per pixel; and, in each pixel, an element (40) for writing the code present at the moment into the memory.