Binocular vision naked eye 3D image generation method

Abstract

The invention discloses a binocular vision naked-eye 3D image generation method, and the method comprises the following steps: carrying out the image collection through a collection device, and obtaining a color + gray original binocular image pair; completing calibration of the acquisition devices for the acquired original binocular image pairs, and acquiring relative positions between the acquisition devices; performing binocular correction on the original binocular image through the relative position between the acquisition devices to obtain corrected binocular images including a black andwhite right-eye image and a color left-eye image; performing binocular matching on the corrected binocular image by using a binocular matching algorithm to generate a parallax image; drawing the blackand white right-eye image by using the parallax image to generate a color right-eye image, synthesizing the color left-eye image and the color right-eye image into one image, and performing pixel interleaving to generate a final binocular image; according to the invention, a binocular matching method with high matching precision is adopted, the hardware cost is reduced, and the method can be applied to a real-time scene with higher precision and comfort.

Description

technical field [0001] The invention relates to the research field of image processing, in particular to a binocular vision naked-eye 3D image generation method. Background technique [0002] With the development of cameras and computer technology, it has become the goal of more and more researchers to let computers learn to understand and interpret images like humans, and the discipline of computer vision was born. Among them, the naked-eye 3D display technology is the front end of the current display technology and the trend of future development. At present, the use of binocular cameras to achieve 3D shooting and production has gradually become the mainstream. Among them, the high-efficiency and low-cost conversion of binocular images into naked-eye 3D display content has become a research hotspot, but the key and difficulty is to achieve high matching accuracy and fast speed. Binocular matching algorithm. [0003] Since it was proposed in the 1970s, naked-eye 3D displa...

AI Technical Summary

Problems solved by technology

[0004] Traditional 2D display technology can only display flat images, but cannot experience the depth of the three-dimensional sense, and cannot satisfy people's pursuit of realistic and stereoscopic vision

However, the application fields of traditional 3D technology are relatively limited, and the fidelity is not enough and the display quality is poor.

Unable to achieve real-time matching, applied to real-time scenarios

The stereoscopic display effects of current 3D videos are uneven, and 3D videos with good stereoscopic display effects often require complex production steps and high production costs, which also lead to the lack of 3D resources and become an important factor restricting the development of naked-eye 3D technology.

The focus and difficulty of binocular stereo vision research has always been how to improve the matching accuracy of binocular matching algorithms. At present, algorithms with relatively high binocular matching accuracy generally have a large amount of calculation, which is very time-consuming and cannot achieve real-time matching. Applied to real-time scenarios

Images