Dynamic Autofocus Search for Multi-Lens Digital Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital image processing apparatuses with multiple lenses are inefficient in autofocus operations due to fixed AF search start positions, leading to repeated and unnecessary movements when focusing on the same object or when lenses are at the center of the search range.
Innovation Solution
The method involves dynamically setting AF search start positions and directions for each lens based on the current focus position, with lenses within the AF safety range setting unique start positions and directions, and those outside the range setting uniform directions and center positions to optimize autofocus efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If fixed AF search start positions are used for multiple lenses, then the AF operation is simple to implement, but the AF efficiency deteriorates when repeatedly focusing on the same object or when lenses are at center position
Solution Approach 1:
The patent applies dynamics by making the AF search start positions and search directions variable rather than fixed. The control unit dynamically adjusts the start positions and directions based on current lens positions and object characteristics, allowing the system to adapt to different focusing scenarios and improve efficiency without excessive complexity
Solution Approach 2:
The patent uses preliminary action by pre-setting multiple possible AF search start positions and directions based on predicted object locations. The system prepares multiple candidate start positions in advance, and selects the appropriate one based on current conditions, thereby reducing the actual search time and improving focusing speed
2Productivity
If multiple lenses perform AF independently at the same fixed positions, then each lens can focus separately, but the system fails to utilize the advantage of having multiple lenses
Solution Approach 1:
The patent applies segmentation by dividing the AF task among multiple lenses with different search start positions and directions. Instead of all lenses performing the same search simultaneously, each lens is assigned a segmented search strategy based on its position and the predicted object location, allowing parallel processing with improved overall efficiency
Solution Approach 2:
The patent uses local quality by assigning different AF search characteristics to different lenses based on their specific positions and the local search requirements. Each lens adapts its search start position and direction to its local context, optimizing the overall system performance rather than using a uniform approach
3Loss of time
If lenses return to fixed AF search start positions before repeated AF, then the AF search range is standardized, but unnecessary lens movements occur increasing operation time
Solution Approach 1:
The patent uses preliminary action by pre-calculating and storing multiple AF search start positions corresponding to different object locations. When an object is detected, the system immediately selects the appropriate pre-prepared start position, eliminating the need for lenses to return to a fixed center position and reducing unnecessary movement time
Solution Approach 2:
The patent applies feedback by using information from previous AF operations and object detection to determine the next AF search parameters. The control unit continuously monitors lens positions and object characteristics, and uses this feedback to dynamically adjust search start positions and directions, maintaining accuracy while reducing operation time
Data Source
AI summary
A method of performing fast autofocus (AF) in a digital image processing apparatus having a plurality of lenses mounted therein. By dynamically setting AF search start positions and AF search directions by using current AF position information, fast AF may be performed.