Traditional optical imaging is based on geometric optics, and uses the principle of "what you see is what you get" for reference, ignoring many high-dimensional optical information. With the traditional optical hardware function approaching the physical limit, it has been unable to meet the application requirements in many fields. However, with the continuous evolution of the new generation of information technologies such as sensors, cloud computing and artificial intelligence, computational imaging, as a new imaging solution, obtains encoded light field information through multiple dimensions, providing a new perception paradigm for sensor design that is far beyond the human eye.

Although computational optical imaging is a new technology, many exciting research results have been achieved, and computational imaging technology has re-examined the imaging process from the perspective of information, and completely changed the way people acquire, process and interpret visual information. For example, in the biomedical field, providing high-resolution, high-contrast microscopic images of cells and tissues for the early and accurate diagnosis of diseases; In industrial manufacturing, help precision parts defect detection and quality control; In terms of security monitoring, achieve high-definition target identification and tracking at a long distance and under a large field of view; In the driverless scenario, the vehicle is provided with accurate environmental perception information to ensure driving safety. These typical applications demonstrate the powerful potential and broad impact of computational optical imaging technology, heralding future innovations and breakthroughs in more fields.