Depth image edge preserving processing method and device

Abstract

The invention relates to a depth image edge preserving processing method and device. The method comprises the steps of an obtaining step: obtaining a depth image and an RGB image in the same scene; a correction step: obtaining a to-be-corrected point in the depth image, obtaining a to-be-processed point corresponding to the to-be-corrected point in the RGB image according to the to-be-corrected point, determining a to-be-processed window of the to-be-processed point in the RGB image by taking the to-be-processed point as a center, and obtaining an absolute value of a pixel difference between a non-to-be-processed point and the to-be-processed point in the to-processed window; according to the absolute value of the pixel difference, obtaining a correction depth value of a to-be-corrected point corresponding to the to-be-processed point in the depth image; and a generation step: generating a corrected depth image of the scene according to the corrected depth value. Through the depth image edge preserving processing method, the depth image to be corrected is clearer, more accurate and smoother.

Description

technical field [0001] The present disclosure relates to the technical field of depth image processing, and in particular, to a method and device for edge-preserving processing of a depth image. Background technique [0002] At present, in the field of dual-camera blurring, whether it is binocular stereoscopic imaging, TOF imaging, or structured light 3D imaging, the obtained depth images may have incorrect data or incomplete data in the occluded area. Therefore, the obtained depth image edge Occurrence of inconsistencies with RGB image edges. figure 1 is a schematic diagram of a depth image without edge-preservation processing, represented by figure 1 It can be seen that the edge of the depth image without edge preservation processing has wrong data, and its depth value on the same plane is not smooth enough. Contents of the invention [0003] In order to overcome the problems existing in the prior art, the present disclosure provides a method and device for edge-preser...

AI Technical Summary

Problems solved by technology

[0002] At present, in the field of dual-camera blurring, whether it is binocular stereo imaging, TOF imaging or structured light 3D imaging, the obtained depth images may have incorrect data or incomplete data in the occluded area. Therefore, the obtained depth image edge Occurrence of inconsistencies with RGB image edges

figure 1 is a schematic diagram of a depth image without edge-preservation processing, which is represented by figure 1 It can be seen that the edge of the depth image without edge preservation processing has wrong data, and its depth value on the same plane is not smooth enough

Images