Method of determining image background, device for determining image background, and a non-transitory medium for same

1. A method of determining an image background, comprising:

recognizing a face region in an image, and obtaining a face distance based on the face region;

obtaining a face distance parameter of each pixel in the image based on the face distance;

processing the face distance parameter and corresponding color parameter of each pixel in the image by applying a pre-trained image region segmentation model to determine an image region type corresponding to each pixel;

determining a background region of the image based on the image region type corresponding to each pixel and performing preset background image processing on the background region;

wherein obtaining the face distance based on the face region comprises:

detecting a width of a central horizontal axis of the face region to obtain a face horizontal axis distance; and/or,

detecting a length of a central vertical axis of the face region to obtain a face vertical axis distance; and

wherein obtaining the face distance parameter of each pixel in the image based on the face distance comprises:

detecting a first distance from each pixel in the image to the central horizontal axis of the face region and calculating a first ratio of the first distance to the face horizontal axis distance;

and/or,

detecting a second distance from each pixel in the image to the central vertical axis of the face region and calculating a second ratio of the second distance to the face vertical axis distance.

2. The method according to claim 1 , further comprising:

recognizing a sample face region of each sample image in an image sample set, and obtaining a sample face distance based on the sample face region;

obtaining a face distance parameter of each pixel in a background region and a face distance parameter of each pixel in a user region in each sample image based on the sample face distance;

training the face distance parameter and corresponding color parameter of each pixel through a back-propagation algorithm by applying a preset neural network to determine neuron parameters on the neural network to obtain the image region segmentation model.

3. The method according to claim 1 , further comprising:

determining a user region of the image based on the image region type corresponding to each pixel and performing preset user image processing on the user region.

4. A terminal device, comprising a memory, a processor and computer programs stored in the memory and executable by the processor, wherein when the computer programs are executed by the processor, the processor is configured to perform acts of:

recognizing a face region in an image, and obtaining a face distance based on the face region;

obtaining a face distance parameter of each pixel in the image based on the face distance;

processing the face distance parameter and corresponding color parameter of each pixel in the image by applying a pre-trained image region segmentation model to determine an image region type corresponding to each pixel;

determining a background region of the image based on the image region type corresponding to each pixel and performing preset background image processing on the background region;

wherein the processor is configured to obtain the face distance based on the face region by acts of:

detecting a width of a central horizontal axis of the face region to obtain a face horizontal axis distance; and/or,

detecting a length of a central vertical axis of the face region to obtain a face vertical axis distance; and

wherein the processor is configured to obtain the face distance parameter of each pixel in the image based on the face distance by acts of:

detecting a first distance from each pixel in the image to the central horizontal axis of the face region and calculating a first ratio of the first distance to the face horizontal axis distance;

and/or,

detecting a second distance from each pixel in the image to the central vertical axis of the face region and calculating a second ratio of the second distance to the face vertical axis distance.

5. The terminal device according to claim 4 , wherein the processor is further configured to perform acts of:

recognizing a sample face region of each sample image in an image sample set, and obtaining a sample face distance based on the sample face region;

obtaining a face distance parameter of each pixel in a background region and a face distance parameter of each pixel in a user region in each sample image based on the sample face distance;

training the face distance parameter and corresponding color parameter of each pixel through a back-propagation algorithm by applying a preset neural network to determine neuron parameters on the neural network to obtain the image region segmentation model.

6. The terminal device according to claim 4 , wherein the processor is further configured to perform acts of:

determining a user region of the image based on the image region type corresponding to each pixel and performing preset user image processing on the user region.

7. A non-transitory computer readable storage medium, having stored therein computer programs that, when executed by a processor, causes the processor to perform a method of determining an image background, the method comprises:

recognizing a face region in an image, and obtaining a face distance based on the face region;

obtaining a face distance parameter of each pixel in the image based on the face distance;

processing the face distance parameter and corresponding color parameter of each pixel in the image by applying a pre-trained image region segmentation model to determine an image region type corresponding to each pixel;

determining a background region of the image based on the image region type corresponding to each pixel and performing preset background image processing on the background region;

wherein obtaining the face distance based on the face region comprises:

detecting a width of a central horizontal axis of the face region to obtain a face horizontal axis distance; and/or,

detecting a length of a central vertical axis of the face region to obtain a face vertical axis distance; and

wherein obtaining the face distance parameter of each pixel in the image based on the face distance comprises:

detecting a first distance from each pixel in the image to the central horizontal axis of the face region and calculating a first ratio of the first distance to the face horizontal axis distance;

and/or,

detecting a second distance from each pixel in the image to the central vertical axis of the face region and calculating a second ratio of the second distance to the face vertical axis distance.

8. The non-transitory computer readable storage medium according to claim 7 , wherein the method further comprises:

recognizing a sample face region of each sample image in an image sample set, and obtaining a sample face distance based on the sample face region;

obtaining a face distance parameter of each pixel in a background region and a face distance parameter of each pixel in a user region in each sample image based on the sample face distance;

training the face distance parameter and corresponding color parameter of each pixel through a back-propagation algorithm by applying a preset neural network to determine neuron parameters on the neural network to obtain the image region segmentation model.

9. The non-transitory computer readable storage medium according to claim 7 , wherein the method further comprises:

determining a user region of the image based on the image region type corresponding to each pixel and performing preset user image processing on the user region.