Camera Sticking Detection Using Pixel Value Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining whether image data from a camera is stuck are inefficient, particularly when luminance values change minimally, making it difficult to detect anomalies, and require high CPU load for pixel value comparisons across all pixels.
Innovation Solution
A sticking determination device and method that uses a CPU, GPU, and memory to compare pixel values of specific target positions across frames, reducing the load by dividing the image data into units and only comparing pixel values at select positions, allowing for earlier detection of sticking without needing to process entire frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pixel values of all pixels are compared across frames to detect sticking, then detection accuracy is improved, but CPU load increases significantly
Solution Approach 1:
The patent divides the image data into multiple divided units (e.g., blocks or regions) and performs sticking determination separately for each unit. This segmentation allows the system to compare pixel values in a distributed manner, reducing the computational burden on the CPU while maintaining detection accuracy across the entire image.
Solution Approach 2:
The patent applies different processing strategies to different parts of the image data. Specifically, it performs detailed pixel value comparisons only in regions where sticking is suspected or where luminance changes are minimal, while using coarser methods in other regions. This local differentiation optimizes the balance between detection accuracy and computational load.
2Speed
If luminance values are used for anomaly detection, then processing speed is improved, but detection accuracy deteriorates when luminance changes minimally
Solution Approach 1:
The patent changes the parameter used for comparison from luminance values to pixel values. Pixel values contain more information than luminance values alone, enabling the system to detect anomalies even when luminance changes are minimal. This parameter change maintains processing efficiency while significantly improving detection accuracy in low-contrast scenarios.
3Reliability
If entire frames are processed for sticking determination, then detection completeness is improved, but processing time increases
Solution Approach 1:
The patent processes image data by dividing it into multiple divided units and performing sticking determination on each unit separately. This approach ensures that the entire frame is covered (maintaining detection completeness) while allowing parallel or sequential processing that reduces overall processing time compared to analyzing the entire frame as a single unit.
Solution Approach 2:
The patent performs sticking determination on divided units rather than requiring complete frame processing before starting analysis. This partial action approach allows the system to begin detection earlier and process multiple units in parallel, reducing total processing time while still achieving comprehensive coverage of the entire frame.
Data Source
AI summary
A sticking determination device includes a CPU, a GPU, a memory, a bus, and an image acquiring unit configured to acquire an image data from the camera and save the image data to the memory. The image acquiring unit transmits, to the CPU, a save completion notification indicative of completion of saving respective one of divided units of the image data. The CPU transmits a read instruction to the GPU to read, from the memory, a pixel value of at least one pixel at a sticking determination target position. The GPU reads the pixel value of the at least one pixel from the memory and transmits the read pixel value to the CPU. The CPU is configured to determine whether the image data is stuck by comparing the pixel value in a current frame with the pixel value in a previous frame.