Rear View Camera Dewarp Fallback for Safe Image Output
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current rear view camera systems lack a comprehensive safety mechanism for image processing, particularly on the main bus and DRAM, leading to potential data effectiveness issues and incomplete signal path safety.
Innovation Solution
The proposed solution involves a rear view camera image processing apparatus with a safety mechanism applied to a dewarp function, utilizing two processing paths: a normal path through non-safety area components and an emergency path solely through safety area components, ensuring distortion correction and effective image delivery even if components in the non-safety area malfunction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the safety mechanism is applied to all components including main bus and DRAM, then the data effectiveness and reliability are improved, but the device complexity and implementation difficulty increase significantly
Solution Approach 1:
The patent segments the image processing system into safety-critical components (ISP, dewarp, S-RVC, LDI) and non-safety components (main bus, DRAM). The safety mechanism is applied only to the essential image processing chain, isolating the safety-critical functions from the complex bus and memory subsystems. This segmentation allows reliability improvement for critical functions without requiring comprehensive safety implementation across all system components.
Solution Approach 2:
The patent applies different quality levels of safety mechanisms to different parts of the system. The safety mechanism is implemented with full rigor in the image processing chain (ISP, dewarp, S-RVC, LDI) where data effectiveness is critical, while the main bus and DRAM operate without the same level of safety mechanisms. This local differentiation optimizes the balance between reliability and complexity by concentrating safety resources where they are most needed.
2Device complexity
If the safety mechanism is applied only to DRAM with ECC-bit, then the implementation complexity is reduced, but the reliability improvement is insufficient due to lack of coverage on main bus and other components
Solution Approach 1:
The patent divides the system into safety-critical image processing components and non-safety support components. By segmenting the safety mechanism application to only the essential image processing chain (ISP, dewarp, S-RVC, LDI), the patent achieves comprehensive reliability improvement for data effectiveness without the need to implement safety mechanisms in all system components including main bus and DRAM.
Solution Approach 2:
The patent performs preliminary safety validation within the image processing chain itself, checking data effectiveness at each processing stage (ISP output, dewarp output, S-RVC output, LDI output). This preliminary action ensures that any corruption or errors are detected and handled within the safety-critical path before affecting the final output, making extensive safety mechanisms in support components unnecessary.
3Manufacturing precision
If the frame memory is transformed to store high-resolution images, then the image quality is improved, but the chip size increases due to required SRAM capacity
Solution Approach 1:
The patent segments the memory system into a small on-chip SRAM for temporary storage of processed image data and an off-chip DRAM for storing high-resolution frame images. The SRAM only needs to hold intermediate processing results, not the entire high-resolution frame, thereby minimizing on-chip memory requirements and keeping chip size small while still supporting high-quality image output.
Solution Approach 2:
The patent moves the high-capacity memory storage from the two-dimensional on-chip SRAM plane to the off-chip DRAM domain, effectively using a different spatial dimension (external memory space) to accommodate high-resolution image data. This allows the chip itself to remain compact while still supporting high-quality imaging through external memory resources.
Data Source
AI summary
Disclosed is a method for processing, by a rear view camera image processing apparatus operated by at least one processor, an original rear image collected by a rear view camera, in which when the rear view camera collects an original rear image, the original rear image is processed through each of a first path and a second path, and an output image generated through the first path is displayed to a display apparatus, and a reduced rear image generated through the second path is temporarily stored.The effectiveness of the output image generated through the first path is judged, and when the output image is not effective, the temporarily stored reduced rear image is generated as a rear output image and delivered to the display apparatus.