Sideway Image Morphing Using Frontal Contour Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image morphing technologies face challenges in efficiently processing human sideway images due to the need for dedicated neural networks to extract sideway contour points, which is time-consuming and results in low precision, leading to poor morphing effects.
Innovation Solution
A method and device for processing images that acquire human contour point information from sideway images, determine the morphing region and orientation, and perform morphing along a specific direction, eliminating the need for dedicated neural networks and reducing overhead, by using existing frontal human contour point information for sideway morphing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a dedicated neural network is trained to extract sideway contour points, then extraction precision is improved, but training time and deployment overhead increase significantly
Solution Approach 1:
The patent applies universality by making the existing frontal contour point extraction neural network perform multiple functions: it extracts both frontal contour points and sideway contour points. The system achieves sideway contour point extraction without training a dedicated neural network, thereby reducing deployment overhead while maintaining extraction precision through the same trained model.
Solution Approach 2:
The patent uses copying by transforming the coordinate system of extracted frontal contour points to generate corresponding sideway contour points. Instead of training a new neural network to directly extract sideway points, the system copies and transforms the results from the frontal extraction process, achieving the same effect with significantly reduced overhead.
2Adaptability or versatility
If a dedicated neural network is trained for sideway contour point extraction, then extraction capability is improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The existing frontal contour point extraction neural network is made universal to handle both frontal and sideway morphing tasks. By adding coordinate transformation logic to the existing system, the patent enables sideway morphing capability without increasing neural network complexity or requiring additional dedicated models.
Solution Approach 2:
The patent introduces coordinate transformation as an intermediary step between frontal contour point extraction and sideway morphing. This intermediary transformation process bridges the gap between the existing frontal extraction capability and the desired sideway morphing function, avoiding the need for complex dedicated neural network training.
3Productivity
If frontal contour point information is used for sideway morphing, then deployment overhead is reduced, but morphing precision may be compromised
Solution Approach 1:
The patent uses coordinate transformation to copy and adapt frontal contour point information for sideway morphing purposes. By mathematically transforming the coordinate system rather than directly using frontal points, the system maintains precision while achieving rapid deployment without dedicated neural network training.
Solution Approach 2:
The patent applies parameter changes by transforming the coordinate system parameters from frontal view to sideway view. This mathematical transformation of spatial parameters enables the use of frontal contour point information for sideway morphing while maintaining the required precision through accurate coordinate conversion.
Data Source
AI summary
A morphing part to be morphed and a morphing effect are acquired. Human body contour point information is acquired by detecting a human contour point of a human sideway image. A morphing region is determined according to the human contour point information and the morphing part. The morphing region is an image region to be morphed in the human sideway image. A human orientation is determined according to the human contour point information. A morphing direction of the morphing region is determined according to the human orientation and the morphing effect. Morphing is performed along the morphing direction in the morphing region.