Image Composition Correction via Face Detection and Aesthetic Rule Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need to improve image quality by enhancing the overall aesthetic effect of images, which existing technologies have not adequately addressed.
Innovation Solution
An image processing method that performs face detection, determines compliance with multi-person composition rules, and crops images based on face position information to improve the overall effect, optionally rotating images to correct scene orientation and displaying composition prompts for incomplete or low-resolution images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If image processing focuses only on detailed effects, then local image quality is improved, but overall aesthetic effect deteriorates
Solution Approach 1:
The patent segments image processing into two distinct modules: one for local detailed effects (face detection, skin smoothing, feature enhancement) and another for overall composition (aesthetic rule evaluation, cropping, rotating). This segmentation allows each module to optimize its specific function without compromising the other, resolving the contradiction between local quality and overall aesthetic effect.
Solution Approach 2:
The patent introduces a new dimension of evaluation by adding composition assessment based on aesthetic rules (rule of thirds, golden ratio, symmetry) to the traditional local detail processing. This dimensional expansion transforms the processing from purely local pixel manipulation to a comprehensive approach that includes spatial arrangement and compositional harmony, thereby improving overall aesthetic effect while maintaining local quality.
2Ease of operation
If image cropping is performed to correct composition, then overall aesthetic effect is improved, but image resolution deteriorates
Solution Approach 1:
The patent applies partial cropping rather than excessive cropping by evaluating multiple candidate crop regions and selecting the one that achieves compositional correction with minimal loss of image content. The system performs partial action by cropping only the necessary portions to correct composition while preserving the maximum possible resolution and image content.
Solution Approach 2:
The patent changes the parameter of crop region selection by evaluating multiple candidate regions with different coordinates and sizes, then selecting the optimal region that balances composition correction with resolution preservation. This parameter optimization allows the system to achieve aesthetic improvement while minimizing resolution loss through intelligent parameter selection.
3Manufacturing precision
If face detection and composition analysis are performed on all images, then image quality is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary face detection and composition analysis on preview images before final image processing. By conducting initial assessment on smaller preview images, the system can quickly identify images requiring detailed processing, thereby reducing overall processing time while maintaining quality improvement for images that need it.
Solution Approach 2:
The patent applies partial processing by performing full composition analysis and aesthetic evaluation only on images that meet certain criteria (e.g., contain multiple faces, show composition issues in preview). For images that already meet aesthetic standards, the system skips intensive processing, thereby reducing processing time while maintaining quality improvement where needed.
Data Source
AI summary
Provided are an image processing method, a device, and a storage medium, the method comprising: performing human face detection on a first image, and obtaining position information for each human face in the first image, then determining whether the first image conforms to a multi-person composition rule on the basis of the position information of each human face in the first image; and cropping the first image based on the position information of each face in the first image to obtain a processed image, in response to the first image not complying with the multi-person composition rule.