The influence of gesture complexity and occlusion problems in 3D hand mesh-based reconstruction work leads to low accuracy of 3D hand network estimation. This paper proposes a novel Hand Feature Occlusion Network (HFONet) designed to address the issue of hand occlusion. The method utilizes CNN for modeling local feature relationships and Transformer for modeling global feature relationships, constructing a feature fusion architecture. Firstly, the original image is inputted into the feature mapping module based on the Connected Feature Pyramid Network. The spatial and channel information of the image is processed through the Connected Feature Pyramid Network, resulting in feature maps of the hand and the object. Next, the two feature maps are inputted into the feature mapping module. By designing different connection methods between different levels, the feature maps are processed using an interactive self-attention mechanism to fully utilize the details and holistic information of the image. Finally, the processed features are inputted into the regressor module to obtain the final 3D mesh model. The results demonstrate that training the Hand Feature Occlusion Network using the H03D dataset yields lower average joint error, mesh error (mm), and F-scores compared to similar methods. Moreover, it achieves higher accuracy in complex and occluded scenes. The proposed structural design in this paper provides higher estimation accuracy in 3D mesh reconstruction.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.