Static Icon Detection in Dynamic Video Using Adaptive Inter-Frame Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting static characteristic information in dynamic images, such as TV station logos, face challenges with varying sizes, shapes, and transparency levels, leading to unsatisfactory processing effects and erroneous estimations due to interference with video backgrounds, and lack automation in detection.
Innovation Solution
A full-automatic detection method and system that compares inter-frame differences in predetermined regions to identify static information pixels, adjusts thresholds as needed, and synthesizes results to accurately detect static characteristic information, including opaque and transparent icons, by using a judgment module to determine static points and a determination module to exclude background information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If inter-frame difference detection is used to detect overlay icons, then detection simplicity is improved, but detection precision deteriorates due to varying icon sizes, shapes, and transparency levels
Solution Approach 1:
The patent applies dynamics by making the detection threshold adaptive rather than fixed. The threshold dynamically adjusts based on local image characteristics and inter-frame difference statistics, allowing the detection system to adapt to varying icon sizes, shapes, and transparency levels while maintaining operational simplicity
Solution Approach 2:
The patent changes the parameter of detection threshold from a static value to a dynamically adjusted parameter. By modifying the threshold based on local variance and inter-frame difference characteristics, the system achieves both simplicity and precision across diverse icon conditions
2Measurement precision
If edge detection method is used to detect icon regions, then detection capability is improved, but automation deteriorates as manual judgment is required
Solution Approach 1:
The detection system performs self-service by automatically determining whether icon regions have been detected and stopping the detection operation autonomously. The system uses statistical criteria based on inter-frame difference accumulation to self-judge detection completion without requiring manual intervention
Solution Approach 2:
The patent implements feedback by continuously monitoring the accumulation of static information points and comparing it against threshold criteria. This feedback mechanism enables the system to automatically determine when detection is complete and stop the operation, achieving full automation
3Stability of the object's composition
If static detection algorithm is applied to video content, then processing consistency is improved, but processing quality deteriorates due to different characteristics of static information
Solution Approach 1:
The patent applies local quality by using different processing approaches for different regions. Static information regions are detected and processed separately from dynamic video content using inter-frame difference analysis, while video content undergoes different processing, thereby achieving both consistency and quality
Data Source
AI summary
According to a method in the present invention, first whether an inter-frame difference for each pixel in predetermined first region and second region in the dynamic image exceeds a predetermined threshold is judged to determine whether the pixel is a static information point, and when, in the second region, pixels in regions not overlapping with the first region are determined as non-static information points, judgment on the inter-frame difference and the predetermined threshold is stopped; and then static characteristic information in the dynamic image is determined based on the static information points in the first region. Preferably, the inter-frame difference for each pixel in the first region may be re-judged based on an adjusted predetermined threshold, to further determine the static characteristic information in the dynamic image, so that static opaque static characteristic information or static characteristic information with arbitrary degrees of transparency in the dynamic image can be detected.