Surveillance Motion Detection Using Non-Linear Weighting and Edge Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance systems face challenges in accurately distinguishing between motion caused by moving objects and illumination changes or clutter motion, leading to false alarms and inefficient background subtraction algorithms, especially in uncontrolled outdoor environments with varying weather conditions.
Innovation Solution
The proposed solution employs non-linear weighting and automatic threshold updating, combined with edge detection using a modified Laplacian of Gaussian filter, to differentiate between pixel differences induced by illumination and those caused by object motion, thereby improving the robustness of background subtraction and reducing false alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a pixel difference threshold is used to differentiate static objects from moving objects, then moving objects can be detected, but false alarms increase due to illumination changes and clutter motion being misclassified as moving objects
Solution Approach 1:
The patent segments the pixel difference image into multiple bands (e.g., low, medium, high difference regions) and applies different processing strategies to each band. This allows the system to handle illumination changes, clutter motion, and true moving objects differently, reducing false alarms while maintaining detection accuracy.
Solution Approach 2:
The patent applies spatially varying thresholds and processing parameters differentially across the image based on local characteristics. For example, different threshold values are used in different spatial regions, and local statistical properties are computed to adapt to regional variations in illumination and clutter patterns.
2Measurement precision
If background image updates are performed frequently to maintain accuracy, then detection precision improves, but system resource consumption and processing complexity increase
Solution Approach 1:
The patent implements periodic background image updates at strategically chosen intervals rather than continuously. The update timing is optimized to balance accuracy requirements with processing load, performing updates when significant scene changes are detected or at predetermined intervals that maximize detection accuracy while minimizing computational overhead.
3Measurement precision
If manual tuning of threshold parameters is performed to optimize detection, then detection accuracy improves, but ease of operation decreases due to requiring expert knowledge and time-consuming calibration
Solution Approach 1:
The patent implements automatic threshold selection and parameter optimization algorithms that adapt to the specific surveillance scene without requiring manual tuning. The system automatically computes optimal thresholds based on local statistical properties of the image data, performs self-calibration during initial operation, and dynamically adjusts parameters in response to changing environmental conditions.
Data Source
AI summary
In some embodiments, a non-transitory processor-readable medium stores code representing instructions to cause a processor to smooth a current image of a scene to produce a smoothed image and subtract pixel values of a background image of the scene from corresponding pixel values of the smoothed image to produce an altitude difference image. Pixel values of the altitude difference image are weighted to produce a weighted difference image. The weighted difference image is convolved to produce a convoluted difference image. A threshold is applied to each pixel of the convoluted difference image to produce a thresholded difference image. Pixels having a value less than the threshold are removed from the thresholded difference image and classified as background pixels. Foreground pixels are determined based on the thresholded difference image.