A kind of image reconstruction method and device

Abstract

The invention provides an image reconstruction method, comprising the following steps of obtaining interval information among pulses in a pulse sequence related to signals on spatial positions, wherein when accumulated strength of spatial position signals exceeds a preset threshold, the pulses are generated, and the pulses are arranged to form the pulse sequence according to a sequence; and on thebasis of the preset threshold and the interval information, determining pixel values of the spatial positions at at least partial moments when image reconstruction is carried out. According to the reconstruction method, a motion process of a high speed motion object can be finely reconstructed, and moreover, in a reconstruction process, through analysis of time interval change among the pulses and establishment of a mapping relationship between the change and reconstruction pixel values of the corresponding spatial positions, the motion object with textures is reconstructed based on tiny processing quantity. In an accumulation mode, sampling data comprises light intensity information at all moments, so according to the reconstruction method provided by the invention, an image at any moment can be reconstructed, and the limitation that only the images at frame sampling moments are reserved in a traditional video is broken.

Description

technical field [0001] The present application relates to the image field, in particular to an image reconstruction method and device, and an image digitization and reconstruction system using the device. Background technique [0002] The form that exists in a certain time and space is the source of information. For example, the propagation of light in space forms a dynamic image, the flow of a large number of water molecules produces ocean information, and the dynamic movement of air molecules and other floating objects forms climate information. As far as dynamic images are concerned, humans and creatures perceive the world by capturing photons through their eyes. Modern cameras use CCD (Charge-coupled Device, charge-coupled device) or CMOS ((Complementary Metal Oxide Semiconductor, Complementary Metal Oxide Semiconductor) to capture photons to record The dynamically changing world produces a large amount of image and video data. [0003] Traditional methods of expressing...

AI Technical Summary

Problems solved by technology

However, such a video frame rate cannot represent faster-changing physical phenomena, so high-frequency cameras have appeared, and their frame rates can reach 1,000 frames per second or even 10,000 frames or even higher. The problem brought about is the large amount of data. As the scale increases, the corresponding collection and processing circuit design is expensive or even impossible. More importantly, the increase in the frame rate means that the exposure time of a single frame is reduced, and the exposure of a single frame image collected is seriously insufficient. The way to make up for it is to increase pixel size, which in turn brings about a reduction in spatial resolution

Images