Method for determining disparity map in binocular virtualization, mobile terminal, and storage medium

Abstract

The invention discloses a method for determining a disparity map in binocular virtualization. The method applied to a mobile terminal comprises: a first camera obtains a first image and a second camera obtains a second image; distortion correction is carried out on the first image and stereo correction is carried out on the first image after distortion correction to obtain a reference image; distortion correction is carried out on the second image and stereo correction is carried out on the second image after distortion correction to obtain a registration image; an optical flow of the registration image relative to the reference image is calculated and a vector angle of the optical flow is calculated based on the optical flow; whether the vector angle of the optical flow exceed a preset angle threshold is determined; and if the vector angle of the optical flow is greater than the preset angle threshold, stereoscopic matching is carried out the reference image and the registration imageto obtain a disparity map of the reference image and the registration image. Besides, the invention also provides a mobile terminal and a computer-readable storage medium. Therefore, the effect of background blurring is improved.

Description

technical field [0001] The invention relates to the field of mobile terminals, in particular to a method for determining a disparity map in binocular blurring, a mobile terminal and a storage medium. Background technique [0002] As dual-camera configurations become increasingly common in mobile terminals such as mobile phones, a binocular image depth estimation method suitable for mobile terminals such as mobile phones becomes very important. [0003] At present, in the process of binocular image depth estimation, stereo matching is required after lens calibration and correction, that is, to estimate the position of each pixel in one image (such as a reference image) in another image (such as a registration image). The corresponding stereo matching (Stereo Matching) relationship, thus the disparity estimation value of each pixel of the image can be obtained, and then a disparity estimation map (DisparityMap) can be obtained. According to the optical principle, the disparit...

AI Technical Summary

Problems solved by technology

If you want to blur the background, but the general DC does not have the function of large aperture, and the point-and-shoot camera cannot achieve the effect of blurring the background

This makes it difficult to realize the blurred background in ordinary mobile phones. Even if some manufacturers adopt it, the effect of blurred background is not good.

Images