Video Object Counting via Multi-Resolution Frame Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for tracking and counting moving objects are computationally expensive due to their reliance on full resolution video images and complex algorithms, which increases processing demands as image resolution and frame rate increase.
Innovation Solution
A method that determines an area of motion by threshold subtracting a current video frame from a short-term and long-term average video scene, using low resolution images created by low pass filtering, and simplifies object detection and tracking by using edge detection and pixel differencing, reducing noise and processing pixels, while maintaining accurate object counting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full resolution video images are used for object tracking and counting, then measurement precision is improved, but device complexity and computational cost increase
Solution Approach 1:
The patent divides the video processing task into multiple stages: first processing at reduced resolution to identify potential objects, then selectively applying full-resolution analysis only to regions containing detected objects. This segmentation approach maintains detection accuracy while significantly reducing overall computational complexity by avoiding full-resolution processing of entire frames.
Solution Approach 2:
The patent applies different processing qualities to different regions of the video stream. Full resolution processing is applied locally only to regions where objects are detected, while other regions use reduced resolution. This local quality approach ensures measurement precision is maintained where needed while reducing device complexity in other areas.
2Productivity
If higher frame rates are used for video processing, then productivity is improved, but use of energy and computational cost increase
Solution Approach 1:
The patent segments the processing workload by first analyzing reduced-resolution frames at high frame rates to detect motion, then selectively processing only frames containing detected objects at full resolution. This allows maintaining high productivity through frequent sampling while reducing energy consumption by avoiding full processing of every frame.
Solution Approach 2:
The patent uses reduced resolution processing for all frames (partial action) and full resolution processing only for frames containing objects of interest. This partial application of full processing maintains productivity by processing many frames while reducing energy consumption compared to processing all frames at full resolution.
Data Source
AI summary
A method and system are provided for counting moving objects in a digital video stream. In contrast to known computationally-expensive methods, areas of motion are determined by threshold subtracting a current video frame from a short term average video scene. An object box surrounding an object is determined by threshold subtracting the current video frame from a long term average video scene. Coordinates of the moving object are identified by associating the area of motion with the object box, if it overlaps the area of motion, to define a moving object box. An event counter can be incremented when the moving object box is in a buffer zone in the current frame, and was in a detection zone in an earlier frame, and was initially detected in a buffer zone on the opposite side of the detection zone.