Achieving practical full projector compensation requires the projection display to adapt quickly to textured projection surfaces and unexpected movements without interrupting the display procedure. A possible solution to achieve this involves using a projector and an RGB camera and correcting both color and geometry by directly capturing and analyzing the projected natural image content, without the need for additional patterns. In this study, we approach full projector compensation as a numerical optimization problem and present a physics-based framework that can handle both geometric calibration and radiometric compensation for a Projector-camera system (Procams), using only a few sampling natural images. Within the framework, we decouple and estimate the Procams' factors, such as the response function of the projector, the correspondence between the projector and camera, and the reflectance of projection surfaces. This approach provides an interpretable and flexible solution to adapt to the changes in geometry and reflectance caused by movements. Benefitting from the physics-based scheme, our method guarantees both accurate color calculation and efficient movement and reflectance estimation. Our experimental results demonstrate that our method surpasses other state-of-the-art end-to-end full projector compensation methods, with superior image quality, reduced computational time, lower memory consumption, greater geometric accuracy, and a more compact network architecture.

Long hours keeping a bad sitting posture can damage one's health, so maintaining a healthy sitting posture is crucial for students who need to sit for a long time. Previous works tended to install sensors around the human body or use cameras to capture the user's posture and use Convolutional Neural Network (CNN) to recognize sitting posture. However, many sensors are expensive and inconvenient to use, and CNN cannot model long-range dependencies. Therefore, we introduced the Vision Transfomer because its internal Self-Attention (SA) structure has the inherent ability to capture long-range dependencies. However, the transformer requires large-scale data for pre-training and ignores potential correlations inter-samples. In addition, the vision transformer only uses a single position embedding, which does not provide adequate position information. To solve these problems, We introduce Convolutional Stems and External-Attention (EA), which allows the model not to require pre-trained and to obtain sample relevance features for the entire dataset. We then propose a new vision transformer structure named Dual-position embedding Transformer (DuFormer), which adds a position embedding before and after the segmented patch, respectively. We later demonstrated the effectiveness of DuFormer through ablation experiments. Using DuFormer, we constructed a model architecture called SPRNet and tested its performance on our dataset. The experiments indicated that SPRNet successfully recognized three habitual sitting postures and outperformed other State-Of-The-Art (SOTA) methods.