Digital Video Noise Reduction via Spatial-Temporal Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise reduction methods in digital moving picture data face challenges in accurately distinguishing between motion effects and noise effects, leading to blurring or loss of motion objects, and inefficiencies in compression and image quality due to noise recognition as a signal component, particularly in video image sequences.
Innovation Solution
A method that combines spatial and temporal filtering, where spatial filtering is applied first to preserve image edges and then temporal filtering is used with adaptive threshold values based on noise energy in the YCbCr color space, reducing the number of frames required for filtering and improving motion detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If simple temporal averaging filtering is used to remove noise, then noise is reduced, but moving pixels become blurred or lost
Solution Approach 1:
The patent applies different filtering strengths to different pixel regions based on local characteristics. Motion-compensated filtering is applied adaptively: stronger filtering to static or slowly varying regions, and weaker or no filtering to fast-moving regions. This local differentiation allows noise reduction in static areas while preserving motion details in dynamic areas.
Solution Approach 2:
The patent introduces motion compensation to make the filtering dynamic rather than static. By estimating motion vectors and compensating for pixel displacement between frames, the filter can track moving objects and apply appropriate filtering only to regions that are truly static or slowly varying, thereby preserving motion clarity while reducing noise.
2Measurement precision
If motion estimation is performed for all pixels to avoid filtering errors, then pixel classification accuracy improves, but computational complexity increases
Solution Approach 1:
The patent segments the image into different regions based on motion characteristics. Rather than performing full motion estimation on all pixels, it identifies and segments fast-moving regions where filtering should be avoided, and applies motion-compensated filtering only to regions where it is beneficial. This segmentation approach reduces unnecessary computation while maintaining classification accuracy.
Solution Approach 2:
The patent performs motion estimation selectively rather than universally. It applies motion compensation only to regions where motion is detected or where filtering is deemed necessary, rather than performing the computationally expensive motion estimation on every pixel in the image. This partial action approach maintains necessary accuracy while reducing overall computational burden.
3Productivity
If noise is compressed along with valid picture data, then compression rate improves, but coding artifacts increase
Solution Approach 1:
The patent performs noise filtering as a preliminary action before compression. By removing noise from the video signal prior to encoding, the subsequent compression process operates on cleaner data with fewer high-frequency components that would otherwise be misinterpreted as signal content. This preliminary noise removal prevents the generation of compression artifacts while maintaining efficient compression ratios.
Data Source
AI summary
Provided is a method of removing noise from digital moving picture data reducing the number of frames used in a temporal filtering operation and able to detect motion between frames easily. The method comprises a method of spatial filtering, a method of temporal filtering, and a method of performing the spatial filtering and the temporal filtering sequentially. The spatial filtering method applies a spatial filtering in a YCbCr color space, preserving a contour/edge in the image in the spatial domain, and generating a weight that is adaptive to the noise for discriminating the contour/edge in the temporal filtering operation. The temporal filtering method applies temporal filtering based on motion detection and scene change detection, compensating for global motion, the motion detection considering the brightness difference and color difference of the pixels compared between frames in the temporal filtering operation, and a weight that is adaptive to the noise for detecting the motion in the temporal filtering operation. The spatial filtering method is preferably performed first, and the temporal filtering method is performed with the result of the spatial filtering.