Compensation method of invalid pixels

Abstract

The invention discloses a compensation method of invalid pixels. The compensation method of invalid pixels includes the steps: according to the positional relation between invalid pixels, performing coarse classification on the invalid pixels, and determining isolate invalid pixels and communication invalid pixels; according to an image classification method, performing coarse classification on remote sensing images to obtain a coarse classification result, and marking the coarse classification result as Class(k); according to the positions of the invalid pixels in the remote sensing images, combining the coarse classification of invalid pixels and the coarse classification of the remote sensing images, performing secondary classification on the invalid pixels, and determining the isolate invalid pixels in a monotonous area, the communication invalid pixels in the monotonous area, the isolate invalid pixels in an edge area and the communication invalid pixels in the edge area; and successively compensating the four types of invalid pixels. The compensation method of invalid pixels realizes compensation correction of invalid pixels of a visible light near infrared area-array COMS (Complementary Metal Oxide Semiconductor) camera, and has the advantages of being high in compensation correction accuracy and guaranteeing imaging quality of images.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a method for compensating invalid pixels. Background technique [0002] Visible and near-infrared area array CMOS (Complementary Metal Oxide Semiconductor) cameras have tens of thousands of pixels, [0003] It is possible to directly obtain the digital image of the reflected sunlight energy of the pixel array target, but due to the limitation of the process, the change of the space environment, the aging of components and other factors during the satellite launch process and post-launch, there is an unavoidable problem in the area array COMS camera. That is, the existence of invalid pixels, the invalid pixels appear in the image as dark spots and bright spots that cannot reflect the real characteristics of the target, if not compensated and corrected, it will seriously affect the quality of the image. [0004] At present, a variety of effective methods have ...

AI Technical Summary

Problems solved by technology

[0005] However, the compensation and correction methods for invalid pixels of visible light and near-infrared CMOS cameras are all considered to compensate invalid pixels from the perspective of spatial processing, and do not consider the difference in radiation response characteristics of different cameras, nor do they consider isolated invalid pixels and connected Therefore, when the existing compensation and correction scheme is used to compensate and correct the invalid pixels, the accuracy of compensation and correction is limited, and the improvement of image quality is limited, which seriously affects the quality of other subsequent quantitative products such as object classification. precision

Images