Automatic Focusing Using Gaussian Weighted Block Definition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing focusing methods based on a region of interest require storing a mask in advance, which takes up storage space.

Innovation Solution

An automatic focusing method that selects a rectangular region of interest on a target image, generates a ROI mask using a Gaussian function with simple parameters, and calculates weight values for each block to prioritize focusing on the region of interest, avoiding the need for pre-stored mask data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a mask of the region of interest is stored in advance, then the region of interest can be focused clearly, but storage space is occupied

Engineering Contradiction:
Improvefocusing precisionVSAvoidstorage space
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies preliminary action by pre-defining the region of interest through user selection before the focusing process begins. The system allows users to select a rectangular region on the screen that indicates where they want to focus, and this selection is stored temporarily in memory (not permanent storage) to guide the subsequent automatic focusing calculation. This resolves the contradiction by preparing the focusing parameters in advance without occupying significant storage space.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses copying by creating a simplified representation of the region of interest as a weight map or mask data structure in memory during the focusing calculation process. Instead of storing actual mask images permanently, the system generates a computational copy (weight values for different regions) that is used temporarily to guide the autofocus algorithm. This temporary in-memory copy enables precise focusing control without requiring permanent storage space.

Inventive Principle:
Principle #26Copying

2Adaptability or versatility

If the region of interest is set by user selection, then flexibility is improved, but system complexity increases

Engineering Contradiction:
Improveregion selection flexibilityVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the image into a grid of blocks and assigning different weight values to different blocks based on their location relative to the user-selected region of interest. The image is segmented into multiple small regions (blocks), and each block is independently weighted. This segmentation approach simplifies the system by breaking down the complex task of regional focusing into manageable discrete units that can be processed independently through straightforward weight assignment and summation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent uses parameter changes by transforming the user's region selection into a set of weight parameters that are applied to different blocks. Instead of implementing complex region-based focusing logic, the system changes the parameter representation from geometric region definitions to numerical weight values. Each block receives a weight parameter (higher for ROI blocks, lower for non-ROI blocks), and the autofocus calculation simply sums these weighted block definitions. This parameter transformation simplifies the system architecture while maintaining flexibility.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3588934B1Automatic focusing method and apparatus based on region of interest
Publication Date: 2022.01.26 SUZHOU KEDA TECH
  • EP3588934B1 patent drawingFigure 1
  • EP3588934B1 patent drawingFigure 2
  • EP3588934B1 patent drawingFigure 3~4B

AI summary

Provided are an automatic focusing method and apparatus based on a region of interest, belonging to the technical field of imaging. The method comprises the following steps: acquiring a target image that has been divided into blocks; acquiring the definition of each block, respectively; acquiring normalized central coordinates and a normalized size of a region of interest on the target image; respectively calculating a full width at half maximum coefficient in the horizontal direction and the vertical direction according to the normalized size; calculating a weight value of each block using a two-dimensional discrete Gaussian function according to the normalized central coordinates and the full width at half maximum coefficient; calculating a normalized overall definition of the target image according to the weight value and definition of each block; and focusing according to the normalized overall definition. The present invention can automatically calculate a mask of the region of interest, thereby avoiding the occupying of storage space required when storing ROI mask data.