Variance Metric Video Feature Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video encoding and transcoding techniques are inefficient in detecting features like scene changes and black borders, relying on ineffective methods such as sound level changes and pixel color counts.
Innovation Solution
The use of sum of variances metrics for pixel blocks to identify video features, such as scene changes, black borders, and complexity, which allows for adaptive transcoding operations like intra-frame encoding, bit allocation, and border region handling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques (sound level change detection, pixel color counting) are used to identify video features, then the transcoding process can be performed, but the detection accuracy and efficiency of video features is poor
Solution Approach 1:
The patent changes the detection parameter from simple pixel color counting or sound level detection to variance-based metric detection. By calculating the variance of pixel values within defined regions (entire frame, border regions, caption regions), the system achieves more accurate feature detection. The variance metric captures subtle changes in pixel distribution that conventional methods miss, thereby improving detection accuracy without significantly increasing computational complexity.
Solution Approach 2:
The patent segments the video frame into multiple regions for differential variance analysis: border regions (top, bottom, left, right edges) and caption regions (typically bottom area). By calculating variance separately for these segments, the system can accurately detect black borders and caption overlays without interference from other frame contents, resolving the contradiction between detection accuracy and processing efficiency.
2Reliability
If accurate video feature detection is implemented, then transcoding quality improves, but the computational complexity and processing time increases
Solution Approach 1:
The patent transforms complex feature detection into simple variance calculations on segmented regions. Instead of using sophisticated image processing algorithms, the system computes variance of pixel values within defined regions and compares against thresholds. This parameter transformation maintains high reliability in feature detection while keeping computational complexity low, as variance calculation is computationally efficient compared to traditional image analysis methods.
Solution Approach 2:
The variance-based detection method is self-adaptive to different video contents. The system automatically adjusts to detect black borders, caption regions, and scene changes without requiring manual configuration or complex machine learning models. The variance metric inherently captures the statistical properties of different video features, allowing the system to maintain high transcoding quality across diverse video types with minimal computational overhead.
3Speed
If conventional feature detection methods are used, then processing speed is maintained, but the detection of black borders and caption regions is ineffective
Solution Approach 1:
The patent segments the frame into border regions and caption regions with predefined geometric boundaries. By calculating variance only within these specific segments, the system efficiently detects black borders and caption overlays. The segmentation approach maintains processing speed by limiting calculations to specific regions rather than analyzing the entire frame, while simultaneously improving detection accuracy through focused regional analysis.
Solution Approach 2:
The patent changes the detection approach from global frame analysis to regional variance analysis. By computing variance metrics specifically for border regions (edges of the frame) and caption regions (typically bottom area), the system achieves rapid detection of these features. The parameter change from global to regional variance calculation maintains processing speed while dramatically improving the accuracy of black border and caption region detection.
Data Source
AI summary
A metric representing the sum of variances for pixel blocks of a region of an image are used to identify the presence a video feature of the image, and a transcoding is performed responsive to identifying the presence of the video feature. The identified video feature can include, but is not limited to, a scene change, the presence of a black border region or a caption region, or the complexity of the image. The transcoding operation can include, but is not limited to, coding the image as an Intra-frame, omitting the content corresponding to the black border region or the caption region from the transcoded image or allocating a relatively lower bit budget for the black border region or a relatively higher bit budget to the caption region during transcoding of the image, or setting the bit budget for rate control during transcoding.