Computational ghost imaging is a technique that uses a single-pixel detector to capture images of objects. This method involves correlating the outputs of two photodetectors: one that measures a field not interacting with the object and another that collects a field interacting with the object. The technique allows for background-free imaging and 3D sectioning, and it demonstrates the classical nature of ghost-image formation. Ghost imaging was initially demonstrated using a biphoton source, leading to interpretations of it as a quantum phenomenon. However, subsequent research showed that ghost imaging can also be performed with pseudothermal light, which is treated as a classical electromagnetic wave. This has sparked interest in the physics of ghost imaging, as it challenges the quantum interpretation. The paper presents a Gaussian-state analysis of ghost imaging that unifies prior work on biphoton and pseudothermal sources. It shows that ghost-image formation is due to classical coherence propagation, with the advantage of high-contrast imagery in the wideband limit. However, other studies claim that pseudothermal-light ghost imaging is fundamentally quantum, involving nonlocal two-photon interference. The paper describes a computational ghost-imaging setup using a single-pixel detector. This configuration does not rely on nonlocal two-photon interference and allows for background-free imaging and 3D sectioning. The setup uses a spatial light modulator (SLM) to create the necessary coherence behavior. The SLM is used to impose a phase on the light, allowing for deterministic modulation of the field. The paper also discusses the use of computational ghost imaging in reflectance and the ability to perform 3D sectioning by precomputing intensity fluctuation patterns. This method allows for the same bucket-detector photocurrent to be correlated with many precomputed intensity patterns, enabling 3D imaging of objects. In conclusion, the paper shows that ghost imaging can be performed with a single-pixel detector by precomputing the intensity fluctuation pattern that would have been seen by a high-resolution detector. This method provides background-free images with controllable resolution and field of view, and it underscores the classical nature of ghost-image formation. The computational ghost imager is a powerful tool for imaging objects with high resolution and the ability to perform 3D sectioning.Computational ghost imaging is a technique that uses a single-pixel detector to capture images of objects. This method involves correlating the outputs of two photodetectors: one that measures a field not interacting with the object and another that collects a field interacting with the object. The technique allows for background-free imaging and 3D sectioning, and it demonstrates the classical nature of ghost-image formation. Ghost imaging was initially demonstrated using a biphoton source, leading to interpretations of it as a quantum phenomenon. However, subsequent research showed that ghost imaging can also be performed with pseudothermal light, which is treated as a classical electromagnetic wave. This has sparked interest in the physics of ghost imaging, as it challenges the quantum interpretation. The paper presents a Gaussian-state analysis of ghost imaging that unifies prior work on biphoton and pseudothermal sources. It shows that ghost-image formation is due to classical coherence propagation, with the advantage of high-contrast imagery in the wideband limit. However, other studies claim that pseudothermal-light ghost imaging is fundamentally quantum, involving nonlocal two-photon interference. The paper describes a computational ghost-imaging setup using a single-pixel detector. This configuration does not rely on nonlocal two-photon interference and allows for background-free imaging and 3D sectioning. The setup uses a spatial light modulator (SLM) to create the necessary coherence behavior. The SLM is used to impose a phase on the light, allowing for deterministic modulation of the field. The paper also discusses the use of computational ghost imaging in reflectance and the ability to perform 3D sectioning by precomputing intensity fluctuation patterns. This method allows for the same bucket-detector photocurrent to be correlated with many precomputed intensity patterns, enabling 3D imaging of objects. In conclusion, the paper shows that ghost imaging can be performed with a single-pixel detector by precomputing the intensity fluctuation pattern that would have been seen by a high-resolution detector. This method provides background-free images with controllable resolution and field of view, and it underscores the classical nature of ghost-image formation. The computational ghost imager is a powerful tool for imaging objects with high resolution and the ability to perform 3D sectioning.