Adaptive Video Noise Reduction via Motion Threshold Switching

The patent applies dynamics by making the noise reduction algorithm adaptive to motion conditions. The system dynamically selects between different processing paths based on detected motion magnitude: using multi-frame alignment algorithms only when motion is small, and switching to single-frame or alternative algorithms when motion is large. This dynamic adaptation resolves the contradiction by ensuring high alignment accuracy is achieved only when motion conditions permit, thereby maintaining processing speed while improving accuracy when possible.

The patent changes the parameter of algorithm selection based on motion detection results. By monitoring motion magnitude as a key parameter and using it to determine which noise reduction algorithm to apply, the system optimizes the balance between alignment accuracy and processing speed. When motion parameters indicate stability, complex alignment algorithms are employed; when motion parameters indicate rapid movement, simpler algorithms are used to maintain real-time processing capability.

The patent segments the noise reduction process into multiple conditional paths based on motion detection. Instead of always applying complex multi-frame alignment algorithms, the system divides processing into different scenarios: low-motion scenarios use full multi-frame alignment for optimal noise reduction, while high-motion scenarios use simplified algorithms. This segmentation reduces overall system complexity by applying complex algorithms only when necessary.

The patent applies partial action by selectively applying multi-frame alignment algorithms only to portions of video sequences where motion conditions are favorable. Rather than uniformly applying complex algorithms to all frames, the system identifies and processes only those frames where motion is sufficiently small, achieving effective noise reduction where possible while avoiding unnecessary computational complexity in high-motion segments.

The patent applies preliminary action by performing motion detection and magnitude assessment before committing to complex multi-frame alignment processing. By evaluating motion conditions in advance, the system can determine whether multi-frame alignment is appropriate, avoiding unnecessary processing time consumption when motion conditions make alignment ineffective. This preliminary assessment prevents wasted computational effort on frames where alignment would fail.

The patent uses dynamics to adaptively adjust processing intensity based on real-time motion conditions. When motion magnitude exceeds thresholds indicating rapid movement, the system dynamically switches from time-consuming multi-frame alignment to faster alternative algorithms. This dynamic adjustment ensures that processing time is optimized by applying computationally intensive methods only when motion conditions justify their use, thereby reducing overall processing time while maintaining noise reduction effectiveness when possible.