Mobile Virtual Clothes Fitting via Face and Shoulder Keypoint Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current virtual clothes-fitting systems on mobile devices face challenges due to limited computation power and complex algorithms, resulting in poor user experience and inefficient interface design, with existing solutions focusing on person detection but not effectively considering how clothes fit on the human body.
Innovation Solution
A method and system for virtual cloth-fitting using video augmented reality on mobile wireless devices, which includes detecting a face and shoulder contour, determining keypoints, mapping clothes onto the user's upper body, and displaying an augmented video stream with adaptive and deformable cloth fitting, reducing processing requirements by focusing on a limited region of interest and eliminating the need for 3D rendering or gender selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If PC-based virtual clothes-fitting systems are used, then clothing fitting accuracy is improved, but device complexity and computation power requirements increase
Solution Approach 1:
The patent extracts and focuses only on the essential components needed for clothes fitting - specifically face detection and shoulder contour detection - rather than implementing full-body 3D scanning. This extraction of key features enables accurate clothes mapping on mobile devices with limited computation power, resolving the contradiction between fitting accuracy and device complexity
Solution Approach 2:
The system segments the body into specific regions of interest (face and shoulder areas) rather than processing the entire body. By dividing the detection task into focused segments, the system achieves accurate clothes fitting while reducing overall computational requirements and system complexity
2Measurement precision
If full-body detection algorithms are implemented, then detection accuracy is improved, but processing speed decreases
Solution Approach 1:
The patent extracts only the necessary detection tasks (face and shoulder contour) from full-body detection, eliminating unnecessary processing steps. This extraction maintains detection accuracy for clothes fitting while significantly improving processing speed by avoiding computation on unrelated body regions
3Difficulty of detecting and measuring
If complex algorithms are used for person detection, then detection capability is improved, but ease of operation deteriorates
Solution Approach 1:
The system performs automatic face and shoulder contour detection without requiring user intervention or manual landmark marking. The algorithm automatically identifies and tracks the necessary body features, providing accurate clothes fitting while maintaining ease of operation - users simply need to position their face and shoulders within the camera view
4Shape
If 3D rendering is implemented, then visualization quality is improved, but computation power requirements increase
Solution Approach 1:
Instead of creating full 3D models and performing complex 3D rendering, the patent uses 2D image copying and mapping techniques. Clothing images are directly mapped onto detected face and shoulder contours in the 2D video stream, achieving satisfactory visualization quality while dramatically reducing computation power requirements for mobile devices
Data Source
AI summary
A system and method for virtual clothes fitting based on video augmented reality in a mobile wireless device is disclosed. In an embodiment, a method for virtual cloth-fitting with video augmentation in a mobile wireless device includes receiving a video stream of an upper body of a user; detecting a face in the video stream; detecting a shoulder contour; determining keypoints of the shoulder and neck portions of the video stream of the upper body of the user according to the detected face and the detected shoulder contour; mapping an image of clothes to the video stream of the upper body of the user according to the keypoints; and displaying an augmented video stream of the upper body of the user with the image of the clothes overlaid over a portion of the video stream of the upper body of the user.