Color Filter Generation Using Pixel Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for generating image filters are time-consuming and require significant storage space, making it difficult to apply specific filter effects like color, brightness, or contrast, as they often include unnecessary texture elements, limiting user convenience.
Innovation Solution
An electronic device with a processor that compares a first image to a second image with a different color pattern to generate a color filter, allowing the filter to be applied to a third image, thereby reducing processing time and storage needs, and enabling efficient generation and application of desired filter effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional style transfer method is used to generate filter, then filter effect can be achieved, but processing time is long and storage space is large
Solution Approach 1:
The patent extracts only the necessary color information from the style image by comparing pixel values between the input image and style image, separating the essential filter effect (color) from unnecessary elements (texture, pattern). This extraction approach eliminates the need for complete style transfer processing, significantly reducing processing time while maintaining filter quality.
Solution Approach 2:
The patent segments the filter generation process into two distinct parts: color information extraction and filter application. By dividing the complex style transfer task into manageable segments (comparing only color channels), the system achieves faster processing without sacrificing the desired filter effect.
2Manufacturing precision
If conventional style transfer method is used to generate filter, then filter effect can be achieved, but storage space requirement is large
Solution Approach 1:
The patent extracts only the essential color data from the style image and discards unnecessary texture and pattern information. This selective extraction dramatically reduces the amount of data that needs to be stored and processed, while still achieving the desired filter effect quality.
Solution Approach 2:
Instead of storing and processing the complete style image as in conventional methods, the patent inverts the approach by generating a compact color filter directly from color value comparisons. This inversion transforms a storage-intensive process into a memory-efficient operation.
3Adaptability or versatility
If complete style image is processed to extract filter, then all style elements are captured, but unnecessary texture is also applied
Solution Approach 1:
The patent extracts only the color component from the style image by comparing pixel values, deliberately leaving out texture and pattern elements. This selective extraction allows users to apply clean color filters without unwanted texture overlays, improving ease of operation while maintaining adaptability for color-based filtering needs.
Solution Approach 2:
The patent applies local quality by treating color information differently from texture information - extracting and preserving color data while discarding texture data. This localized processing approach ensures that only the desired color filter effect is applied, enhancing user convenience for color-based filtering applications.
Data Source
AI summary
According to an embodiment of the present disclosure, an electronic device comprises: a camera; a display; a memory; and at least one processor electrically connected to the display and the memory. The processor obtains an input of selecting a first image, generates a color filter by comparing the first image and a second image having a color pattern different from the first image, obtains a third image, and generates a fourth image by applying the generate color filter to the third image. The processor derives a function by comparing color values of pixels of the first image and the second image, and coefficients of the function. The color filter may include the function derived by comparing the first image and the second image, the coefficients of the function, and a color lookup table generated on the basis of the derived function. The processor stores the generated color filter in the memory and generates the fourth image by applying the stored color filter to the third image.