Phase Unwrapping Confidence-Based Rework

The patent applies preliminary action by pre-calculating and storing confidence factors for each node before the actual phase unwrapping process. These confidence factors represent the reliability of phase differences between neighboring nodes. During unwrapping, the algorithm simply retrieves these pre-computed confidence factors instead of performing complex error analysis in real-time, significantly reducing computational intensity and improving processing speed.

The patent segments the phase unwrapping process into two distinct phases: (1) a pre-processing phase where confidence factors are calculated and stored for all nodes, and (2) an execution phase where unwrapping is performed using the pre-computed confidence factors. This segmentation allows computationally intensive operations to be performed once during pre-processing, while the actual unwrapping operation becomes computationally lightweight and fast.

The patent implements feedback by using confidence factors to guide the unwrapping process. When a closed loop is detected during unwrapping, the algorithm uses the pre-computed confidence factors to identify which nodes are likely erroneous and requires re-processing. This feedback mechanism enables automatic error correction without increasing algorithmic complexity, as the confidence factors provide direct guidance on where corrections are needed.

The patent replaces the traditional mechanical approach of detecting and correcting phase errors during unwrapping with a confidence-factor-based system. Instead of performing complex error detection and correction operations during the unwrapping process itself, the system uses pre-computed confidence factors to predict and guide corrections, substituting a simpler, more reliable mechanism for the traditional approach.

The patent applies preliminary action by pre-calculating and storing confidence factors for each node before the actual phase unwrapping process. These confidence factors represent the reliability of phase differences between neighboring nodes. During unwrapping, the algorithm simply retrieves these pre-computed confidence factors instead of performing complex error analysis in real-time, significantly reducing computational intensity and improving processing speed.

The patent applies dynamics by adaptively adjusting the unwrapping process based on confidence factor thresholds. When confidence factors indicate high reliability, the algorithm proceeds quickly without additional verification. When confidence factors fall below thresholds, the algorithm dynamically triggers re-processing of affected nodes. This dynamic adaptation optimizes processing time by avoiding unnecessary computations in high-confidence regions while focusing computational resources on uncertain regions.