In-Vehicle Camera ECU Processing Load Reduction via Frame Alternation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In-vehicle cameras face increased processing loads due to the need to generate specific image data like bird's eye-view and reduced-screen images, which burdens the camera Electronic Control Unit (ECU) with additional processing requirements.
Innovation Solution
The imaging apparatus performs different processing on sequential frames and adds headers to image data, allowing the ECU to identify and selectively process frames without needing to generate these specific images, thereby reducing processing load by alternating or randomly applying different processing to frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the camera ECU performs processing to obtain particular image data such as bird's eye-view conversion image data and reduced-screen image data, then the required image data types are achieved, but the processing load of the camera ECU increases
Solution Approach 1:
The patent divides the processing of different image data types across multiple sequential frames captured by the image sensor. Instead of generating all required image variants (visual image, bird's eye-view, reduced-screen) from a single frame using heavy ECU processing, the system captures multiple frames and applies different processing to each frame at the sensor level, segmenting the processing workload temporally and spatially.
Solution Approach 2:
The patent performs image processing operations in advance within the image sensor itself rather than later in the ECU. By alternating processing methods across sequential frames at the sensor level, the system prepares multiple variants of image data beforehand, eliminating the need for the ECU to perform heavy conversion operations later.
2Ease of operation
If the camera ECU processes all frames with the same processing method, then processing is simplified, but the variety of output image data types is limited
Solution Approach 1:
The patent implements periodic alternation of processing methods across sequential frames. The image sensor alternates between different processing operations (e.g., visual image processing, bird's eye-view conversion, reduced-screen processing) for each captured frame, enabling diverse output types while maintaining a systematic and manageable processing approach.
Solution Approach 2:
The patent introduces dynamic processing where the processing method applied to each frame varies based on a predetermined pattern or random selection. This allows the system to adaptively generate different image data types from the same input sequence, transforming a static single-processing approach into a dynamic multi-processing system.
Data Source
AI summary
An imaging apparatus includes: an image sensor; a signal processing unit repeatedly performing processing different from each other in a given number of sequential frames to imaging signals obtained by the image sensor to thereby obtain image data of respective frame; and an image data output unit sequentially outputting the image data of respective frames.