Video Feature Detection Reuse for Computational Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Methods for processing video data, such as object detection or recognition, are computationally intensive due to the need to process redundant information in video frames where portions of the video remain unchanged over time.
Innovation Solution
A method that identifies unchanged portions of video frames and reuses feature data from previous frames to reduce processing demands, using a convolutional neural network (CNN) to generate and combine feature data from both frames, thereby reducing the computational load and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video frames are processed using feature detection operations, then accurate feature detection is achieved, but computational load and power consumption increase
Solution Approach 1:
The video processing is segmented into two distinct paths: a first processing path for frames with motion (where full feature detection is performed) and a second processing path for static frames (where feature detection is skipped or simplified). This segmentation allows the system to process only the necessary portions of each frame, reducing overall computational load and power consumption while maintaining feature detection accuracy when needed.
2Loss of information
If feature detection operations are performed on all video frames, then complete feature data is obtained, but processing time and computational resources increase
Solution Approach 1:
The system performs a preliminary determination of whether motion is present in a video frame before committing to full feature detection processing. By evaluating motion presence first, the system can avoid unnecessary feature detection operations on static frames, significantly reducing processing time while ensuring that feature data is fully processed only when motion is detected.
3Productivity
If motion detection is performed before feature detection, then processing efficiency is improved, but additional processing steps are required
Solution Approach 1:
The motion detection and feature detection operations are merged into a unified processing framework where the motion detection serves as a gating mechanism for the feature detection. Rather than treating them as completely separate operations, the system integrates them so that the outcome of motion detection directly controls whether feature detection is performed, reducing overall processing steps while maintaining efficiency.
Data Source
AI summary
A method of processing video data representative of a video comprising a first and second frame to generate output data representative of at least one feature of the second frame. The method includes identifying a first and second portion of the second frame, which correspond to a first and second portion of the first frame, respectively. First feature data obtained by processing first frame data associated with the first portion of the first frame using a first feature detection operation is retrieved from storage. Second feature data representative of a second feature map is generated by processing second frame data associated with the second portion of the second frame using the first feature detection operation. The first feature data and the second feature data are processed using a second feature detection operation to generate the output data.