Optical Tracking Device Removing Moving Background Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional optical object tracking devices fail to effectively eliminate noise caused by a moving background, leading to inaccurate object tracking, as they rely on differential image calculations which are insufficient when the ambient light source position changes.
Innovation Solution
The device captures a background image when the light source is off and an operating image when it's on, then uses a processing unit to calculate a differential image, identify objects, and remove noise by comparing background and object information, using size and distance thresholds to distinguish valid objects from background noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If differential image calculation is used to remove ambient light image, then fixed ambient light noise is eliminated, but moving background noise cannot be effectively eliminated
Solution Approach 1:
The patent segments the image processing into multiple stages: first capturing background images at different times, then calculating differential images between consecutive background images to identify and remove moving background noise, and finally calculating the differential image between the operating image and the processed background image. This multi-stage segmentation allows the system to distinguish between fixed and moving background elements.
Solution Approach 2:
The patent performs preliminary action by capturing background images at multiple time points before the actual object tracking. By pre-capturing and processing background images to identify and remove moving noise, the system prepares a cleaned background model that can be used during subsequent object tracking operations, ensuring that moving background noise does not interfere with object detection.
2Device complexity
If only differential image between bright and dark frames is calculated, then processing is simple, but moving background noise remains in the image
Solution Approach 1:
The patent divides the image processing into distinct segments: capturing background images at multiple time points, calculating differential images between consecutive background images to identify moving noise, and then using this processed information to remove noise from the final differential image. This segmentation increases processing complexity but significantly improves tracking accuracy by addressing moving background noise.
3Measurement precision
If background images are captured and processed to remove moving noise, then object tracking accuracy improves, but processing time increases
Solution Approach 1:
The patent performs background image capture and processing as a preliminary action before object tracking begins. By pre-processing background images to identify and remove moving noise, the system prepares a cleaned background model that can be efficiently used during subsequent object tracking operations, reducing the processing time required during actual tracking while maintaining high accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively eliminates moving background noise, improving the accuracy of object tracking and making the device suitable for use on movable devices like vehicles or portable electronics.
Implementation Method 1
The light source is configured to emit light at a lighting frequency
Implementation Method 2
the image sensor is configured to capture a plurality of image frames containing the object
Data Source
AI summary
An optical tracking device includes a light source, an image sensor and a processing unit. The light source emits light at a lighting frequency. The image sensor outputs an operating image when the light source is being turned on and outputs a background image when the light source is being turned off. The processing unit is configured to obtain background information from the background image, calculate a differential image of the operating image and the background image, obtain object information from the differential image and compare the background information and the object information thereby removing background noise.