Object Tracking via Background Subtraction and Intensity Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current object tracking technologies face challenges in detecting and monitoring fast-moving or obscured objects, such as UAVs, due to issues with visibility and resolution, leading to difficulties in timely identification and response.
Innovation Solution
An object tracking system that processes a sequence of images to estimate pixel background values, subtract background, compute summed intensity values, cluster identified values, and determine the location and velocity of moving objects using shift-and-add processes and image stabilization techniques, enabling real-time detection and classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional object tracking algorithms are used to detect fast-moving objects, then the tracking process can handle basic motion, but the detection accuracy deteriorates and false positives increase for fast-moving or obscured objects
Solution Approach 1:
The system performs preliminary background subtraction and motion prediction before final detection, preparing processed data in advance to handle fast-moving objects more effectively
Solution Approach 2:
The system transitions from analyzing single frames to processing sequences of frames with temporal dimension, using frame-to-frame correlation to improve detection of fast-moving objects by tracking their motion trajectory over time
2Reliability
If conventional object tracking algorithms are used, then basic tracking can be performed, but visibility of obscured objects deteriorates leading to missed detections
Solution Approach 1:
The system performs preliminary background subtraction to remove static elements before detecting moving objects, enhancing the visibility of obscured objects by eliminating background clutter
Solution Approach 2:
The system continuously processes sequences of frames rather than isolated images, maintaining continuous tracking of objects through occlusion by predicting their trajectories and maintaining detection across multiple frames
3Measurement precision
If detailed analysis of each video frame is performed, then detection accuracy can be improved, but processing time increases making real-time tracking difficult
Solution Approach 1:
The system extracts only the essential features (motion vectors, intensity changes, background differences) from full video frames, discarding redundant information while preserving detection accuracy
Solution Approach 2:
The system performs preliminary processing steps (background subtraction, motion estimation) on simplified data before detailed analysis, reducing the computational burden of frame-by-frame examination
4Speed
If motion models are employed to handle fast-moving objects, then trajectory prediction can be improved, but system complexity increases
Solution Approach 1:
The system adjusts motion model parameters (velocity, acceleration, prediction horizon) dynamically based on object characteristics and scene conditions, handling fast-moving objects by optimizing parameters for their specific motion patterns
Data Source
AI summary
Systems and methods for tracking moving objects in accordance with embodiments of the invention are disclosed. In one embodiment of the invention, an object tracking system comprises a processor, a communications interface, and a memory configured to store an object tracking application. The object tracking application configures the processor to receive a sequence of images; estimate and subtract background pixel values from pixels in a sequence of images; compute sets of summed intensity values for different per frame pixel offsets from a sequence of images; identify summed intensity values from a set of summed intensity values exceeding a threshold; cluster identified summed intensity values exceeding the threshold corresponding to single moving objects; and identify a location of at least one moving object in an image based on at least one summed intensity value cluster.