Video Recording System Noise Removal with Motion-Adaptive Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video recording systems face challenges in accurately removing random noise, which can cause flickering and lower picture quality, and often result in afterimages or forged colors, especially when processing digital video signals with motion detection, leading to increased processing load and circuit scale.
Innovation Solution
A video recording system comprising a converter unit for orthogonal transformation, a correlation produce unit, a filtering unit controlled by a controller unit, and signal processing means that produce color signals, brightness, and color difference signals to accurately discriminate between motion and still portions, reducing noise and processing load while avoiding forged colors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If frame circulating type noise removing apparatus is used to remove random noise, then noise removal effect is improved, but afterimage occurs at motion portions
Solution Approach 1:
The patent applies different circulation coefficients to different spatial frequency components and motion regions. Specifically, it uses motion detection to identify motion portions and applies a first circulation coefficient to still portions and a second circulation coefficient (lower than the first) to motion portions, thereby preventing afterimage while maintaining noise removal effectiveness in different regions
Solution Approach 2:
The patent dynamically adjusts the circulation coefficient based on motion detection results and spatial frequency characteristics. The circulation coefficient is not fixed but varies according to the detected motion state and spatial frequency band, allowing the system to adapt to different video content and prevent afterimage in motion portions while maintaining noise removal in still portions
2Object-generated harmful factors
If circulation coefficient is reduced at motion portions to prevent afterimage, then afterimage is suppressed, but noise removal efficiency decreases
Solution Approach 1:
The patent segments the video signal into different spatial frequency components using Hadamard transformation and processes each component with appropriate circulation coefficients. It also segments the image into motion and still portions using motion detection, applying different processing parameters to each segment, thereby maintaining noise removal efficiency in still portions while preventing afterimage in motion portions
Solution Approach 2:
The patent applies different circulation coefficients to different spatial frequency components and motion regions. Specifically, it uses motion detection to identify motion portions and applies a first circulation coefficient to still portions and a second circulation coefficient (lower than the first) to motion portions, thereby preventing afterimage while maintaining noise removal effectiveness in different regions
3Measurement precision
If Hadamard transformation is used to detect motion, then motion detection accuracy is improved, but device complexity increases
Solution Approach 1:
The patent uses Hadamard transformation to generate transformed video signals that copy the essential motion information in a compressed form. The transformation coefficients capture motion characteristics without requiring full pixel-by-pixel comparison, thereby improving motion detection accuracy while reducing the computational complexity and circuit scale compared to traditional methods
Data Source
AI summary
A video camera comprises an optic system 101, an image pickup element 102, a de-mosaicing (de-tessellating) process portion 103, frequency converter portions 104-106, a frame memory 124, parameter produce portions 125-127, noise reduction process portions 107-109, a frame memory 110, frequency converter portions 111-113, noise reduction process portions 114-116, frequency inverter portions 117-119, a brightness signal produce portion 120, a color difference produce portion 121, a coding process portion 122 and a recording medium 123. With the above-mentioned structures, it is possible to execute noise extraction depending on the characteristics of an input video signal, and also to obtain an effect of improving high S/N, but without producing deterioration of the picture quality, i.e., removing the noises of the picked up video, effectively.