A multi-line time-lapse scanning extended sampling sub-pixel image registration method

Abstract

The invention provides a method for registering sub-pixel images through multi-linear-array time difference scanning expansion sampling. The method includes the steps that (1) a multi-linear-array time difference scanning detection device is constructed; (2) a plurality of linear array detectors conduct imaging on the same position in a view field in sequence, and the step (3) is conducted immediately after imaging is completed; or each linear array detector conducts scanning imaging in the scanning direction according to a sampling interval corresponding to the sampling frequency set in the step (1), Nt groups of image data are obtained through each time of imaging, and the step (3) is conducted immediately; (3) the Nt groups of image data of each linear array detector are processed to form a frame detection image; (4) the frame detection images obtained by the linear array detectors correspondingly are spliced to obtain the two sub-pixel images; (5) non-uniformity correction is conducted on the two sub-pixel images respectively; (6) the sub-pixel image which is formed firstly is used as a standard, the sub-pixel image which is formed later is moved forwards by Lp lines in the scanning direction; (7) the affine transformation coefficients of the two sub-pixel images are obtained, and matching of the two sub-pixel images is completed through the coefficients.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to a multi-line time-difference scanning extended sampling sub-pixel image registration method. Background technique [0002] At present, there are various ways to improve the image resolution, specifically the following ways: improving the image receiver CCD, reducing the pixel size, reducing the pixel pitch, increasing the number of pixels, etc., but due to the current It is difficult to break through due to the limited improvement technology and the influence of quantum efficiency; using software to directly perform difference on a single image does not increase the amount of information in the difference image, so it does not essentially improve the resolution of the image; using Microlens technology and pupil resolution technology will cause the disadvantages of excessive system volume; in general, the existing technology is difficult to overcome the disadvantages of high cost, la...

AI Technical Summary

Problems solved by technology

[0002] At present, there are various ways to improve the image resolution, specifically the following ways: improving the image receiver CCD, reducing the pixel size, reducing the pixel pitch, increasing the number of pixels, etc., but due to the current It is difficult to break through due to the limited improvement technology and the influence of quantum efficiency; using software to directly perform difference on a single image does not increase the amount of information in the difference image, so it does not essentially improve the resolution of the image; using Microlens technology and pupil resolution technology will cause the disadvantages of excessive system volume; in general, the existing technology is difficult to overcome the disadvantages of high cost, large volume, and difficult implementation, which cannot meet the needs of further development in some occasions. need

For inter-frame registration technology, due to the influence of detector cross-pixels and noise, even if the registration technology is very mature, it will be affected by the low resolution of single-frame images, resulting in insufficient registration accuracy, and ultimately making the detection target accuracy Therefore, it is very necessary to seek a technology that improves the resolution of a single frame image and cooperates with a suitable registration method to improve the accuracy of later detection targets.

Images