Image Processing Sequence Generation With Preliminary Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing sequence generation techniques, such as those using genetic programming and genetic algorithms, face inefficiencies in specifying a desired image processing sequence due to prolonged processing times, especially with increased filter types or gene lengths, leading to suboptimal performance.
Innovation Solution
A method and device for generating image processing sequences that involve creating a sequence set, comparing output images to targets, calculating evaluation values, maintaining sequences that meet similarity criteria, and registering reference sequences, iteratively refining the process until a sequence satisfying a reference condition is specified, utilizing genetic programming and image processing layers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If genetic programming is used to generate image processing sequences by repeatedly evaluating individuals over many generations, then an image processing sequence satisfying a reference condition can be specified, but the processing time becomes longer
Solution Approach 1:
The patent calculates a temporary evaluation value as an estimate before performing actual image processing. This preliminary estimation allows the system to filter out sequences that are unlikely to satisfy the reference condition, thereby reducing the number of full evaluations needed and decreasing processing time while maintaining reliability.
Solution Approach 2:
The patent uses the temporary evaluation value as feedback to guide the selection of sequences for full evaluation. By incorporating this intermediate assessment into the evolutionary process, the system can more efficiently converge on satisfactory sequences without requiring excessive generations, thus reducing processing time while ensuring reference condition satisfaction.
2Adaptability or versatility
If the types of image filters used in the image processing sequence increase or the gene length becomes longer, then more complex image processing can be achieved, but the processing time until specifying a satisfactory sequence becomes longer
Solution Approach 1:
The patent applies partial action by using temporary evaluation values that assess only the most critical aspects of sequence performance before full evaluation. This allows the system to handle more complex sequences with multiple filter types and longer gene lengths without proportionally increasing processing time, as the preliminary assessment filters out clearly unsuitable candidates early in the evaluation process.
3Ease of manufacture
If image processing is executed using a registered filter column for each image, then image processing can be performed, but the processing time until specifying a satisfactory sequence becomes longer
Solution Approach 1:
The patent performs preliminary image processing using registered filter columns to generate temporary evaluation values before final sequence selection. This preliminary action allows the system to efficiently evaluate sequence candidates using pre-processed image data, reducing the time required for subsequent full evaluations and accelerating the overall sequence specification process.
Data Source
AI summary
In a method for generating an image processing sequence, when a next routine sequence set is used as a sequence set in a routine of this time, a reference sequence stored in a storage device is compared with each of a plurality of image processing sequences forming a sequence set, and an output image according to the image processing sequence is generated, using a processed image associated with the reference sequence, with respect to a sequence that is the same as the reference sequence and that is formed by an image input layer and a sequence element coupled continuously from the image input layer, of the image processing sequences.