Video Object Detection via Background Subtraction and A/R Algorithm
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting abandoned and removed objects in video streams are ineffective, especially in complex and crowded environments.
Innovation Solution
A method involving background subtraction and an A/R algorithm to analyze video frames, determining the presence of static objects by comparing current frames with background images, and storing the status of these objects without using additional information from the background subtraction process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If background subtraction is performed to detect static objects, then detection capability is improved, but false alarms increase in complex and crowded environments
Solution Approach 1:
The patent segments the detection process into multiple independent stages: background subtraction to identify potential static objects, region analysis to evaluate spatial characteristics, and A/R algorithm application to determine object status. This segmentation allows each stage to focus on specific aspects, reducing false alarms while maintaining detection capability in complex environments.
Solution Approach 2:
The patent introduces an intermediary A/R algorithm that acts as a mediator between the background subtraction results and the final detection output. This intermediary process uses the static region mask and current frame information to determine whether detected objects are truly abandoned or removed, filtering out false alarms while preserving genuine detections.
2Productivity
If traditional detection methods are used in complex environments, then processing speed is maintained, but detection accuracy decreases
Solution Approach 1:
The patent performs preliminary background subtraction and static region identification before applying the A/R algorithm. This preliminary action prepares the data in advance, allowing the main detection process to proceed efficiently without requiring complex real-time analysis, thus maintaining processing speed while improving accuracy in complex environments.
Solution Approach 2:
The patent creates a background model copy that represents the expected scene without moving objects. By comparing the current frame against this copied background model using the A/R algorithm, the system can efficiently detect changes without reprocessing the entire scene, maintaining speed while improving detection accuracy through the copied reference framework.
Data Source
AI summary
A method for processing a time-ordered sequence of video frames. The method is implemented by execution of program code on a processor of a computer system. Each frame includes a two-dimensional array of pixels and a frame-dependent color intensity at each pixel. A current frame and at least one frame occurring prior to the current frame in the sequence are analyzed via a background subtraction on the at least one frame to determine a background image and a static region mask associated with a static region. The background subtraction determines an existence of a static object relating to the static region. A status of the static object is determined, the status being either that the static object is an abandoned object or that the static object is a removed object. The determined status is stored in a data storage medium of the computer system.