Multifunctional three-dimensional shooting method based on multiple-lens shooting

Abstract

The invention mainly discloses a multifunctional three-dimensional shooting method based on multiple-lens shooting. The multifunctional three-dimensional shooting method includes: 1 performing synchronous focusing on a main camera and an auxiliary camera; 2 performing image caching on the main camera and the auxiliary camera; 3 calculating the cached images to obtain a depth diagram; 4 producing three-dimensional effects by combining the depth diagram with a high-definition image of the main camera; 5 enabling users to adjust the three-dimensional effects; and 6 outputting the three-dimensional images. The multifunctional three-dimensional shooting method is free of limitation from spaces, effectively avoids serious three-dimensional errors of the images, is applied to portable devices like intelligent cellphones, and provides the portable devices with high-quality computer vision functions and large expansibility.

Description

technical field [0001] The invention relates to image processing technology, in particular to a multi-purpose stereo camera method based on multi-lens shooting. Background technique [0002] Various camera systems based on multi-lens have appeared in the market at present, and the main purpose of these systems is to perform 3D shooting. However, the current multi-camera systems generally have a series of problems such as inconspicuous 3D effects or serious overlapping of shooting results, and poor processing effects on distant or close-up shots. These problems are particularly prominent on small shooting devices such as 3D smart phones and card cameras. [0003] The above-mentioned problems are mainly due to the fact that these dual-camera cameras directly capture the left and right images, and then directly overlap the left and right images to obtain stereoscopic images. This direct shooting method has many disadvantages: 1) First, the positions of the left and right lens...

AI Technical Summary

Problems solved by technology

This direct shooting method has many disadvantages: 1) First, the positions of the left and right lenses are fixed on this small camera device, which has strong limitations and cannot be used for different shooting scenarios, such as putting on the device screen, viewing position and other parameters make adjustments

2) The current stereo camera assumes that the two stereo cameras are parallel to the epipolar line when shooting and superimposing images, and the two cameras are on the same plane

But it is very difficult to achieve the epipolar line and place the two cameras on the same plane in actual shooting

Especially in the actual stereoscopic shooting, the shaking of the operator's hand and the deviation of the machine calibration are extremely likely to cause small position differences.

In the final stage of actual imaging, due to this kind of disparity, only one pixel offset may cause serious stereo errors in the image

Images