Embedded Frame Counters for Frozen Image Detection in Video Codecs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Functional safety in systems, particularly in the automotive sector, is compromised by the 'frozen image' hazard, where obsolete image content is reused instead of updated versions, leading to potential safety issues in video acquisition, processing, and display systems, especially in systems using compressed full frame buffering and video codecs like H.264 and H.265.
Innovation Solution
A method and system that embeds frame counter information using standard-compliant tools, allowing for end-to-end lossless frame counter transmission and detection, preventing frozen image events by using external blocks for detection and prevention, applicable to both uncompressed and compressed full frame buffers, without adding overhead to the bit stream, and employing watermarking techniques to ensure correct recovery of frame counter information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If compressed full frame buffering is used in video systems, then video transmission efficiency is improved, but the system becomes vulnerable to frozen image hazards
Solution Approach 1:
The patent embeds frame counter information in advance within the video data stream before transmission. This preliminary action enables the receiving end to verify frame updates without adding transmission overhead, thus maintaining video transmission efficiency while preventing frozen image hazards through proactive safety checks
Solution Approach 2:
The patent introduces frame counter information as an intermediary element that mediates between the video data stream and the functional safety requirement. This intermediary enables safety verification without disrupting the efficient compressed video transmission, resolving the contradiction between productivity and reliability
2Reliability
If frame counter information is embedded in video data stream, then frozen image detection capability is improved, but bit stream overhead increases
Solution Approach 1:
The patent embeds frame counter information locally within specific regions of the video data stream rather than adding global overhead. This localized approach enables frozen image detection while minimizing bit stream overhead by only adding necessary safety information where needed
Solution Approach 2:
The patent changes the representation parameter of frame information from traditional metadata to embedded frame counters within the video data stream. This parameter change enables detection capability while maintaining bit stream efficiency by using the existing video data structure
3Ease of manufacture
If standard-compliant video codec IPs are reused, then development cost is reduced, but frozen image protection capability is lost
Solution Approach 1:
The patent makes standard-compliant video codec IPs multi-functional by enabling them to simultaneously perform video compression and frame counter embedding. This universality allows reuse of existing codecs while adding frozen image protection capability, thus reducing development cost without sacrificing safety
Solution Approach 2:
The patent enables standard video codec IPs to self-serve the functional safety requirement by automatically embedding frame counter information during normal video processing. This self-service approach allows cost-effective reuse of existing codecs while gaining protection capability without additional dedicated safety hardware
4Measurement precision
If lossless compression is applied to preserve frame counter information, then detection accuracy is improved, but compression efficiency decreases
Solution Approach 1:
The patent applies lossless compression locally only to specific regions containing frame counter information rather than the entire video stream. This localized lossless compression preserves frame counter accuracy while maintaining overall compression efficiency by allowing lossy compression in other video regions
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
A method, adapted for use in countering e.g. so-called "frozen image" events, includes: - providing (11) a set of data such as e.g. frame data, - composing (13) a data stream including compressed data (153) from the set of data having embedded therein (16) a respective counter which identifies the set of data by lossless encoding the respective counter in the compressed data, - transmitting (13, 30, 21) the composed data stream over a transmission channel (30), - receiving the transmitted composed data stream and recovering therefrom (26) the lossless encoded counter.