Structured light depth camera self-correction method and device for smart phone

Abstract

The invention discloses a structured light depth camera self-correction method and device for a smart phone. The device is composed of an infrared laser speckle projector, an image receiving sensor, aself-correction module, a depth calculation module and a mobile phone application processing AP. According to the device and the method, a speckle pattern is projected by a projector; a feature blockis set in the reference speckle image, an input speckle image is acquired through an image receiving sensor; an optimal matching block corresponding to the feature block is searched in the input speckle image through a similarity criterion; the offset between the feature block and the matching block is obtained; once the optical axis of the projector and the optical axis of the image sensor change relatively. the offset can be changed along with the offset; and an optimal offset os solved according to a certain rule and the reference speckle image is reversely adjusted, so that the center ofthe input speckle image and the center of the reference speckle image can form a self-feedback adjustment closed-loop system, thereby realizing that the input speckle image and the corrected referencespeckle image can always find an optimal matching relation when an optical axis is changed in a relatively large range.

Description

technical field [0001] The invention relates to the technical fields of smart phones, smart cameras, and depth perception, in particular to a self-calibration method and device for a structured light depth camera of a smart phone. Background technique [0002] Depth camera is a camera with depth detection function, which has a wide range of applications and development prospects in 3D information extraction, object recognition, classification, motion tracking, machine vision, human-computer interaction, 3D printing, virtual reality and other fields. Embedded applications of depth cameras on smartphones will also gradually become popular. [0003] The structured light depth camera is mainly composed of a laser coded pattern projector, an image receiving sensor and a depth perception computing module. It can use monocular structured light or binocular structured light to obtain more accurate depth information without being affected by ambient light. The perception algorithm a...

AI Technical Summary

Problems solved by technology

In the self-calibration process, the mobile phone application processing AP is used to set the size and number of feature blocks, match the search window range, and find the best offset. The entire depth-sensing self-calibration process can also be completed by software acceleration or a dedicated IP core. In order to solve the problem of

Images