Achieving true magnification in parallel ghost imaging at zero cost based on the cone beam characteristics of the X-ray tube

Краткое изложение: Ghost imaging (GI), as a novel imaging technique, facilitates image acquisition under low light conditions through single pixel measurements, thus holding great potential in various application areas ranging from biomedical imaging, remote sensing imaging, biometrics, astronomy to 3D imaging. However, to reconstruct high resolution images, GI typically requires a large number of single pixel samplings, which is extremely time consuming and poses practical limitations to its applications. Parallel ghost imaging treats each pixel of the position sensitive detector as a bucket detector and simultaneously performs tens of thousands of ghost imaging in parallel. In previous work, we gradually achieved parallel ghost imaging with high pixel resolution, low dose, and ultra large field of view. Parallel ghost imaging has demonstrated excellent performance and great potential. All this is so exciting. But since all our experiments were carried out at synchrotron radiation facilities, with a series of almost luxurious conditions such as nearly infinite and continuous light supply time, monochromatic, pure, and energy adjustable X rays, expensive and precise experimental equipment, and complete supporting facilities, many peers lacking experimental conditions cannot replicate parallel ghost imaging. Meanwhile, the high cost also hinders its cross field integration. Furthermore, we got rid of the synchrotron radiation source and completed the pipeline style acquisition of parallel ghost imaging in a way that uses rough and inexpensive equipment and is most imitable by others. We achieved high quality ghost imaging with an effective pixel size of 8.03 \textmu m and an image size of 2880×2280 at a laboratory X ray source. The total cost of transforming an X ray computed tomography device into a parallel ghost imaging experimental platform is only \$40. Parallel ghost imaging has been generalized from synchrotron radiation sources to X ray tubes.

However, a key problem remains unsolved. The object arm signal on our laboratory light source was obtained through artificial fitting, and the true magnification relationship between the reference arm and the object arm has not been established. In synchrotron radiation, we achieved true magnification using different magnifying optical lens groups. On the one hand, such a set of lenses is very expensive, making the generalization of parallel ghost imaging difficult again. On the other hand, the flux of the X ray tube is very small, which leads to extremely low efficiency. In this work, we find that compared with the parallel beam of synchrotron radiation, the cone beam of the X ray tube naturally has the characteristic of true magnification by gradually moving the detector away from the light outlet. We only use one detector. When collecting the object arm signal, the detector is moved to a position 30 cm away from the light outlet, and when collecting the reference arm signal, the detector is moved to a position 150 cm away from the light outlet. These two positions form a true magnification relationship of 5 times, achieving super resolution of parallel ghost imaging on the X ray tube. A series of high quality ghost imaging results with an effective pixel size of 7.095 \textmu m and an image size of 2880×2280 in pipeline style acquisition were obtained. The realization of true magnification based on the X ray tube is a prerequisite for achieving ultra large field of view and low dose imaging. Completing this work at zero cost implies great application value and commercial potential.