Method and system for generating wide-dynamic-range irradiance image

Abstract

The invention relates to the field of image processing and discloses a method and a system for generating a wide-dynamic-range irradiance image. The method comprises the following steps of: binarizing each image in an image sequence; selecting one of the binarized images as a reference image, and registering other binarized images with the reference image respectively; and synthesizing the wide-dynamic-range irradiance image according to each registered image. In the binarized images, each pixel is either 0 or 1 to ensure that the difference between non-reference images and the reference image is easy to calculate, the registration of the other images and the reference image can be realized in a mode of low computation complexity and lower resource consumption respectively, and the wide-dynamic-range irradiance image can be generated on handheld video equipment.

Description

technical field [0001] The invention relates to the field of image processing, in particular to the generation technology of high dynamic range radiation map in the field of image processing. Background technique [0002] Dynamic range represents the range from the "darkest" brightness to the "brightest" brightness contained in an image. The larger the dynamic range, the richer the layers that can be expressed, and the wider the color space it contains. In real scenes, especially in some scenes with different lighting conditions, the dynamic range is very large and can contain five or more orders of magnitude of brightness. However, the existing Charge Coupled Device (CCD for short) and Complementary Metal-Oxide Semiconductor (CMOS for short) sensors can only capture 256-1024 brightness levels. Therefore, when using a CCD or CMOS sensor to shoot a scene with a high dynamic range, part of the brightness level information will be lost, either by losing the details of the low...

AI Technical Summary

Problems solved by technology

[0007] However, the inventors of the present invention found that the solution has too high requirements on the sensor equipment, and it is not conducive to the integration with the traditional camera system, so it has no practical value

Although the second option only uses traditional sensors, the optical flow field-based algorithm used in the registration process requires a large number of computing units and storage space, which is costly and slow, and can only be run on a general-purpose computer, which is not conducive to Hardware Implementation in a Handheld Camera Device

[0008] That is to say, regardless of scheme 1 or scheme 2, it is impossible to register images with different exposure times on a handheld camera device and shoot scenes with high dynamic range

Images