Image Sensor On-Chip Occlusion Detection Using Reduced Frame Comparisons
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors face challenges in detecting occlusions, such as dust or water, which can impede light from reaching the sensor, leading to obstructed views in critical applications like autonomous driving or rear-view cameras, without providing immediate user awareness of the obstruction.
Innovation Solution
A system and method for detecting occlusions by generating reduced representations of images, comparing temporally adjacent frames for differences indicative of occlusions, and verifying their presence, with the capability to provide notifications to users through visual, audible, or haptic means, enabling continuous and low-power occlusion detection on-chip.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional image sensors are used without occlusion detection, then device complexity is low, but reliability deteriorates due to undetected occlusions affecting image quality
Solution Approach 1:
The image sensor array is divided into multiple blocks, with specific blocks designated as reference blocks for occlusion detection. This segmentation allows the system to use certain regions as references to detect occlusions in other regions, improving reliability without requiring complete system redundancy.
Solution Approach 2:
The patent creates a reference image from reference blocks that copies the structural information of the scene. This reference image is then compared with images from non-reference blocks to detect occlusions. The copying approach allows occlusion detection without requiring additional physical sensors, maintaining low device complexity.
2Reliability
If continuous occlusion detection is implemented, then reliability improves through constant monitoring, but use of energy increases due to continuous processing
Solution Approach 1:
The occlusion detection system operates by periodically comparing reference images with current images at regular intervals. This periodic action allows continuous monitoring capability while enabling the system to enter low-power states between comparison cycles, reducing overall energy consumption compared to truly continuous processing.
Solution Approach 2:
The system performs occlusion detection only on specific non-reference blocks using reference blocks as comparison targets, rather than processing the entire image array continuously. This partial action approach reduces the computational load and energy consumption while maintaining effective occlusion detection coverage.
3Speed
If reduced representations are used for occlusion detection, then processing speed improves, but measurement precision deteriorates due to data reduction
Solution Approach 1:
The image data is segmented into reference blocks and non-reference blocks, with occlusion detection performed on specific segments rather than the entire image. This segmentation allows for faster processing of individual blocks while maintaining detection precision through targeted comparison with corresponding reference segments.
Solution Approach 2:
The patent extracts only the necessary comparison data from reference blocks and non-reference blocks to perform occlusion detection. By extracting and comparing only the relevant block data rather than processing complete high-resolution images, the system achieves faster processing speed while maintaining sufficient detection precision through selective data extraction.
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 high-speed, continuous occlusion detection, enhancing safety and convenience by alerting users to potential obstructions and maintaining image sensor functionality, even in real-time applications.
Implementation Method 1
The image sensor includes an array of pixels having photosensitive elements (e.g., photodiodes) that absorb a portion of the incident image light and generate image charge upon absorption of the image light
Data Source
AI summary
An imaging system including an image sensor coupled to a controller to image an external scene is described. The controller includes logic storing instructions that when executed causes the imaging system to perform operations including capturing images, including a first image and a second image, of an external scene, and generating reduced representations of the images including a first reduced representation associated with the first image and a second reduced representation associated with the second image. The operations further include comparing the first reduced representation with the second reduced representation to determine a difference between the first image and the second image and identifying an occurrence of an occlusion affecting the image sensor imaging the external scene when the difference is greater than a threshold value.