Mono-pixel camera video imaging system based on compression sensing theory

Abstract

The invention provides a mono-pixel camera video imaging system based on compression sensing theory, which is mainly composed of a light source module, a DMD light modulating module, a signal collecting module, and an algorithm recovery module. Particularly, the system takes a traditional mono-pixel camera as basis, and adopts a multi-spectrum light source to irradiate a target object, thus the light reflected by the target object is gathered to the DMD; after the light signal is modulated by the DMD, the light is gathered to the light sensing face of the detector through a collimating lens and an optical fiber, and converted to light intensity numerical value through an A / D converter, thus the light is matched with an observing matrix to introduce to a TVAL3 algorithm, and the system canform image; due to the limit of the instrument performance parameter and the software algorithm, the imaging speed of a traditional mono-pixel camera is slow and cannot reach the requirement of videoimaging. Through modulating a collecting card parameter and deeply optimizing DAQ, the mono-pixel camera video imaging system improves the imaging quality and realizes the video imaging in a multi-spectrum scale.

Description

technical field [0001] The invention belongs to the field of optical imaging systems and image processing, in particular to a single-pixel camera system. Background technique [0002] With the rapid development of science and technology today, while the amount of information and data that people need to obtain is increasing exponentially, it also poses great challenges to traditional data acquisition technologies. In terms of imaging technology, such as MRI and digital cameras, the COMOS array sensor on the digital camera collects a large amount of data, and then compresses and discards 90% of the data. In the process of collection, a large amount of data is collected only for subsequent deletion, so why not discard a large proportion of redundant information at the beginning, only collect a small part of data, and decompress a large amount of original data from these small amounts of data? Related information. [0003] In 2006, a new type of original signal or image was p...

AI Technical Summary

Problems solved by technology

However, the scheme does not address the sampling method

If the traditional sampling method is adopted, since it is an integrated acquisition assistant, the relevant algorithm structure is cumbersome, and feedback is required for each sampling, which greatly increases the continuous sampling period and reduces the sampling rate. Sampled at 10 frames per second

Images