UDAC: Under-Display Array Cameras

• Published in: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) pp. 1077 - 1084
• Main Authors: Wang, Chengyu, Li, Jing, Madhusudanarao, Pavan C., Hu, Jinhan, Singh, Jitesh K., Choi, WooJhon, Lee, Seok-Jun, Sheikh, Hamid R.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 17-06-2024
• Subjects: Cameras; Image color analysis; Image quality; Image restoration; Organic light emitting diodes; Pattern recognition; Pipelines
• ISSN: 2160-7516
• DOI: 10.1109/CVPRW63382.2024.00114

Under-display cameras (UDC) provide uninterrupted display experience on smartphones, but the captured images suffer from quality degradations such as diffraction blur, and reduced signal-to-noise ratio due to the presence of the OLED panel which obstructs the camera. In this work, we propose to use under-display array cameras (UDAC) consisting of two cameras placed under the OLED display panel to further reduce image artifacts and enhance image quality. We rotate the OLED panel by 45 degrees for one of the cameras to utilize the complimentary nature of the point spread function. We also propose an algorithm to combine images from the two cameras, which includes occlusion and optical flow analysis, image warping, color match and image blending. Experiments on real UDC devices show that the proposed method is superior in reducing the artifacts and improving the signal-to-noise ratio. The proposed method is simple but effective, and it can be easily extended to various configurations and incorporated with various algorithms.