Depth-of-field image acquisition method and device

Abstract

The invention provides a depth-of-field image acquisition method and device. The method comprises the following steps: acquiring a single target image; and constructing a neural network, wherein the neural network is used for carrying out feature extraction and fusion on the target image for multiple times so as to obtain a depth-of-field image of the target image. By using the method, a single image can be obtained based on a common camera; a multi-scale and multi-level feature extractor is arranged in a neural network to extract a feature image of each layer; and then the plurality of feature images are fused to obtain a multi-scale and multi-level depth-of-field image, thereby facilitating a user to perform three-dimensional modeling or simulation by using the depth-of-field image, andfurther providing convenience for the user to perform complex three-dimensional image processing based on a single image. Meanwhile, the equipment cost is greatly reduced in a mode of processing the single image to obtain the depth-of-field image.

Description

technical field [0001] The invention relates to the field of computers, in particular to a method and device for acquiring a depth map. Background technique [0002] With the development of science and technology, users have higher and higher requirements for photos. For example, physical technology can be used to improve the resolution, improve the color contrast and other means to enhance the user's perception. [0003] In the prior art, the depth of field information is often acquired through a special camera, a binocular camera, or a laser ranging device, and the cost of the device is relatively high. In addition, the current related algorithms for constructing depth-of-field images have the problem of a single feature extraction structure, which makes the extracted feature information relatively limited, which is not conducive to the construction of complex 3D images. Contents of the invention [0004] The embodiment of the present invention provides a method and de...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, the depth of field information is often acquired through special cameras, binocular cameras, or laser ranging equipment, and the equipment costs are relatively high.

In addition, the current related algorithms for constructing depth-of-field images have the problem of a single feature extraction structure, which makes the extracted feature information relatively limited, which is not conducive to the construction of complex 3D images.

Images