Litcius/Paper detail

AF-Net: A Convolutional Neural Network Approach to Phase Detection Autofocus

Chi‐Jui Ho, Chin‐Cheng Chan, Homer H. Chen

2020IEEE Transactions on Image Processing16 citationsDOI

Abstract

It is important for an autofocus system to accurately and quickly find the in-focus lens position so that sharp images can be captured without human intervention. Phase detectors have been embedded in image sensors to improve the performance of autofocus; however, the phase shift estimation between the left and right phase images is sensitive to noise. In this paper, we propose a robust model based on convolutional neural network to address this issue. Our model includes four convolutional layers to extract feature maps from the phase images and a fully-connected network to determine the lens movement. The final lens position error of our model is five times smaller than that of a state-of-the-art statistical PDAF method. Furthermore, our model works consistently well for all initial lens positions. All these results verify the robustness of our model.

Topics & Concepts

AutofocusConvolutional neural networkArtificial intelligenceRobustness (evolution)Computer scienceComputer visionFocus (optics)Lens (geology)Feature extractionFeature (linguistics)Pattern recognition (psychology)OpticsPhysicsGeneChemistryLinguisticsPhilosophyBiochemistryImage Processing Techniques and ApplicationsAdvanced Vision and ImagingOptical measurement and interference techniques