Human body noise reduction method and system for multiple tof cameras based on downtilt direction

Abstract

The invention discloses a human body noise reduction method for multiple TOF cameras based on a downward inclination direction, comprising acquiring three-dimensional point cloud data of human bodies of multiple TOP cameras; mapping all three-dimensional point cloud data of human bodies into two-dimensional images; All two-dimensional images are subjected to morphological filtering to obtain filtered two-dimensional images; the filtered two-dimensional images are subjected to background difference processing to obtain depth images with only human bodies; endpoint recognition processing is performed on each depth image with only human bodies, Each endpoint is obtained; through each endpoint, the posture recognition of the human body is performed to obtain a specific probability classification, and then the noise reduction of the human body in the depth image is realized. The invention solves the problems of high complexity and time-consuming 3D human posture recognition algorithm; it solves the problems of low detection accuracy, high noise, low pixel resolution and limited field of view of a single TOF. Combining depth information in angle space or larger space.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method and system for reducing human body noise of multiple TOF cameras based on a downward tilt direction. Background technique [0002] In the prior art, the working principle of the TOF camera is as follows figure 1 As shown, TOF is the abbreviation of Time of Flight, literally translated as the meaning of flight time. The so-called time-of-flight 3D imaging is to continuously send light pulses to the target, then use the sensor to receive the light returned from the object, and obtain the distance information of the target object by detecting the flight (round-trip) time of the light pulse, which is the depth image (hereinafter referred to as the depth image) ). [0003] At present, human body detection through TOF cameras is also an important research direction in the field of computer vision. At present, most methods for human body posture detection are...

AI Technical Summary

Problems solved by technology

[0006] Furthermore, it is necessary to maintain a high accuracy rate of gesture recognition and reduce the complexity of the algorithm to realize that the embedded technology is not yet mature, and the algorithm complexity is high, resulting in more time-consuming. For detection scenarios that require high frame rates also does not apply;

[0007] At present, a single TOF has inherent shortcomings such as low detection accuracy, high noise, and low pixel resolution, and there are limitations in the shooting angle and field of view of a single TOF. For the combination of depth information in 360-degree space or a larger space Together there is currently no complete solution

Images