Depth image denoising method, foreground segmentation method and human motion monitoring method

Abstract

The invention discloses a depth image denoising method, a foreground segmentation method and a human motion monitoring method. The method comprises the following steps: partitioning a depth image, marking a first feature point and a second feature point of each block based on a base plane, calculating a first parameter carrying isolated information and spatial distribution information, clusteringeach block by representing attribute parameters of each block as a feature set, further filtering basin blocks; and reserving the plateau block and judging the difference between the gray levels of the first and second feature points of the nearest extension block of the plateau area to perform edge protection to complete denoising. And for the denoised ROI, foreground segmentation is carried outby using a multi-level contour line. For a segmented human body object, motion monitoring is realized by performing plane fitting on a segmented part and calculating an included angle between a fitting plane and a basic plane. Modeling or a large number of iterative operations are not needed, the operation process is simple, the processing efficiency is high, and the accuracy is high.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a depth image denoising method, a foreground segmentation method and a human body motion monitoring method. Background technique [0002] Depth images can reflect the depth information of the research object, and have good data support for the study of moving objects. [0003] Depth image acquisition tools (such as Kinect sensors) usually calculate depth images by emitting and receiving infrared reflections of infrared sensors in space. However, if the measured object is too far away from the target or multiple reflections occur at the target, then the research object There will be a lot of uneven noise around. In the depth image, the noise also has depth information, and the noise is irregular and uncertain. Therefore, the noise has a great influence on the subsequent analysis and calculation of the depth information of the research object. influences. Most directly, usually th...

AI Technical Summary

Problems solved by technology

However, these methods require a large amount of data processing work such as modeling and iterative calculations, the methods are relatively complex, the efficiency is low, and the denoising effect is not ideal.

Images