Video Frame Analysis for Computational Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video analytics systems face high computational burdens due to repetitive processing of video frames, which can lead to increased power consumption and reduced quality when techniques like downsizing or lowering frame rate are used to mitigate this, without effectively addressing the computational intensity.
Innovation Solution
Implementing a dual computation engine system where critical frames are processed by a computationally intensive AI engine and non-critical frames by a less intensive heuristic engine, using motion vectors to classify frames and determine processing paths, thereby reducing overall computational load without sacrificing analytics quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video frames are processed using a computationally intensive AI system, then analytics quality is maintained, but computational burden and power consumption increase
Solution Approach 1:
The patent segments the video stream into critical and non-critical frames based on motion analysis. Critical frames (with high motion vectors indicating significant changes) are processed by the computationally intensive AI system, while non-critical frames are processed by a simpler, lower-power system. This segmentation resolves the contradiction by applying high computational resources only where necessary for maintaining analytics quality.
Solution Approach 2:
The patent applies different processing qualities to different portions of the video stream based on local characteristics (motion content). Frames with significant motion changes receive high-quality AI processing, while frames with minimal changes receive reduced-quality processing. This local quality approach maintains overall analytics quality while reducing average power consumption.
2Use of energy by moving object
If video frames are downsized or frame rate is lowered to reduce computational burden, then power consumption decreases, but analytics quality deteriorates
Solution Approach 1:
The patent dynamically adjusts the processing approach for each frame based on its characteristics. Instead of uniformly downsizing or lowering frame rate, the system analyzes motion vectors and adapts the processing intensity per frame. This dynamic approach prevents quality deterioration by maintaining full resolution and frame rate for critical frames while reducing processing only for non-critical frames.
Solution Approach 2:
The patent changes processing parameters (computational intensity, processing depth) based on frame characteristics rather than uniformly changing physical parameters like resolution or frame rate. Motion vector analysis drives parameter changes in processing strategy, allowing quality preservation while reducing computational burden where appropriate.
3Measurement precision
If all video frames are processed with high computational intensity, then analytics quality is maintained, but device complexity and processing requirements increase
Solution Approach 1:
The patent segments processing requirements based on frame characteristics. A motion analysis component identifies critical frames requiring intensive AI processing, while non-critical frames are handled by simpler processing paths. This segmentation reduces device complexity by avoiding the need to process all frames with maximum computational intensity.
Solution Approach 2:
The patent introduces motion vector analysis as an intermediary component that mediates between the video input and the AI processing system. This intermediary analyzes frame characteristics and directs only necessary frames to the complex AI system, reducing overall processing requirements while maintaining analytics quality for critical events.
Data Source
AI summary
Techniques are provided for processing video frames in a process flow that includes first and second computation engines. In an example, the first engine is an artificial intelligence based computation engine, and the second engine is a heuristics-based computation engine. A sequence of frames of a video includes a first and second frames that are two consecutive frames in the sequence. An analyzer determines whether the second frame has non-redundant information relative to the first frame. In response to the determination, the analyzer selects one of the first or second engine for processing at least a section of the second frame. For example, if the second frame has non-redundant information relative to the first frame, at least the section of the second frame is processed by the first engine. If the second frame does not include non-redundant information, the second frame is processed by the second engine.