Structured light dynamic scene depth acquiring method based on encoding network template

Abstract

The invention discloses a structured light dynamic scene depth acquiring method based on an encoding network template. The structured light dynamic scene depth acquiring method mainly solves the problems that the depth measurement precision is low and the measurement effects are liable to be affected by a measurement scene in the prior art. The structured light dynamic scene depth acquiring method is implemented through the steps of: designing a binary network template encoded by a binary cubic De Bruijn sequence; respectively acquiring a reference planar graph and a deformed grid image through the binary network template; performing line extraction on the deformed grid image and the reference planar graph to obtain a broken line graph and a vertical line graph of the grid image as well as a broken line graph and the vertical line graph of the reference planar graph; decoding the vertical line graphs of the deformed grid image and the deformed grid image respectively to obtain corresponding vertical line serial numbers; acquiring a point on the reference planar graph matched with a pixel point on a deformed grid image according to the vertical line serial numbers; and calculating a depth value of each point according to a triangulation ranging principle. The structured light dynamic scene depth acquiring method based on the encoding network template is high in measuring accuracy, has good robustness for complex scenes, and can be used for measuring dynamic three-dimensional scenes.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a method for acquiring the depth of a dynamic scene with structured light, which can be used for three-dimensional reconstruction or target recognition. Background technique [0002] With the development of computer vision and industrial level, traditional two-dimensional images can no longer meet the needs of people's production and life. Depth information, which reflects the three-dimensional information of the object's surface, greatly improves people's ability to recognize and understand the complexity of real-world objects. Depth information is widely used in the medical field, cultural relic reconstruction, automatic navigation, human-computer interaction and other fields. How to quickly and efficiently obtain the depth information in the 3D scene from the scene has become a current research hotspot. [0003] According to whether contact measurement needs to be p...

AI Technical Summary

Problems solved by technology

[0013] The purpose of the present invention is to address the deficiencies of the prior art, and propose a structured light dynamic scene depth acquisition method based on coded grid templates, which can reduce measurement errors and improve dynamic scene depth without increasing equipment complexity and calculation complexity. Measurement accuracy of scene depth values

Images