Abstract
Single-pixel imaging (SPI) can reduce the cost and have the potential of being competent for some challenging tasks. However, for a SPI system, acquiring detailed information from a complex scene for complex vision task is a measurements-consuming process, by which low efficiency resulted is one of the important obstructions of SPI for practical application. Reasonable allocation of resources of measurements and calculations is one of the effective solutions to this problem. As a preprocessing procedure with a role of guidance, salient object detection can help the vision system focus more attention on the area with more important information to improve the efficiency. Therefore, in this letter, we explore the implement of salient object detection based on SPI system and present a scheme via discrete cosine spectrum (DCS) acquisition and deep learning model.
Highlights
A S A computational imaging system, Single-pixel imaging (SPI) can break through the limit of hardware by powerful data postprocessing [1]
For a SPI system, acquiring detailed information from a complex scene to meet the needs of some complex vision tasks always consumes a great many measurements and results low efficiency, which is one of the important causes to block the practical process of SPI
We explore the implement of salient object detection based on a SPI system and present a scheme via more efficient acquisition of discrete cosine spectrum (DCS) and deep learning model
Summary
A S A computational imaging system, SPI can break through the limit of hardware by powerful data postprocessing [1]. Such an ability can be introduced into a SPI system as a preprocessing step to locate the areas with more important information in the scene, which can direct the SPI system to focus more detection and calculation on these areas to improve the efficiency of subsequent vision tasks. Simulation and experiment are conducted to verify feasibility of the proposed scheme
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
