Method of processing image, electronic device, storage medium, and program product
The present disclosure provides a method of processing an image, an electronic device, and a storage medium, which may be used in a field of artificial intelligence, especially in a field of image processing, etc. The method includes: acquiring an input image containing a plurality of rows of pixels; performing, by using a plurality of dedicated processing units, a pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in the input image, so as to obtain row data for each row of pixels; and stitching the row data for each row of pixels, so as to obtain an output image.
1. A method of processing an image, comprising:
acquiring an input image containing a plurality of rows of pixels;
performing, by using a plurality of dedicated processing units, a pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in the input image, obtaining row data for each row of pixels; and
stitching the row data for each row of pixels, obtaining an output image, wherein
the input image contains at least one channel, and each of the at least one channel comprises a plurality of rows of pixels;
the performing a pixel extraction in parallel comprises:
padding on an edge of at least one side of the input image, obtaining a padded input image; and
performing, by using the plurality of dedicated processing units, the pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in each channel of the padded input image, obtaining row processing data for each row of pixels in the channel; and
the stitching row data for each row of pixels comprises:
stitching the row processing data for each row of pixels in the channel, obtaining channel data for the channel; and
stitching the channel data for each channel, obtaining the output image.
2. The method of claim 1 , wherein the performing a pixel extraction in parallel further comprises:
determining, based on a width of a convolution kernel used in performing the pixel extraction, a number of pixels being extracted each time the pixel extraction is performed on each row of pixels by each of the plurality of dedicated processing units.
3. The method of claim 1 , wherein the performing a pixel extraction in parallel further comprises:
determining, based on a stride in a width direction of a convolution kernel used in performing the pixel extraction, a pixel being extracted each time the pixel extraction is performed on each row of pixels by each of the plurality of dedicated processing units.
4. The method of claim 1 , wherein the performing a pixel extraction in parallel further comprises:
determining, based on a stride in a height direction of a convolution kernel used in performing the pixel extraction, an order of performing the pixel extraction on the plurality of rows of pixels by each of the plurality of dedicated processing units.
5. An electronic device, comprising:
at least one processor; and
a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions, when executed by the at least one processor, cause the at least one processor to implement operations of processing an image, comprising:
acquiring an input image containing a plurality of rows of pixels;
performing, by using a plurality of dedicated processing units, a pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in the input image, obtaining row data for each row of pixels; and
stitching the row data for each row of pixels, obtaining an output image, wherein
the input image contains at least one channel, and each of the at least one channel comprises a plurality of rows of pixels;
the performing a pixel extraction in parallel comprises:
padding on an edge of at least one side of the input image, obtaining a padded input image; and
performing, by using the plurality of dedicated processing units, the pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in each channel of the padded input image, obtaining row processing data for each row of pixels in the channel; and
the stitching row data for each row of pixels comprises:
stitching the row processing data for each row of pixels in the channel, obtaining channel data for the channel; and
stitching the channel data for each channel, obtaining the output image.
6. The electronic device of claim 5 , wherein the instructions, when executed by the at least one processor, cause the at least one processor further to implement operation of:
determining, based on a width of a convolution kernel used in performing the pixel extraction, a number of pixels being extracted each time the pixel extraction is performed on each row of pixels by each of the plurality of dedicated processing units.
7. The electronic device of claim 5 , wherein the instructions, when executed by the at least one processor, cause the at least one processor further to implement operation of:
determining, based on a stride in a width direction of a convolution kernel used in performing the pixel extraction, a pixel being extracted each time the pixel extraction is performed on each row of pixels by each of the plurality of dedicated processing units.
8. The electronic device of claim 5 , wherein the instructions, when executed by the at least one processor, cause the at least one processor further to implement operation of:
determining, based on a stride in a height direction of a convolution kernel used in performing the pixel extraction, an order of performing the pixel extraction on the plurality of rows of pixels by each of the plurality of dedicated processing units.
9. A non-transitory computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions allow a computer to implement operations of processing an image, comprising:
acquiring an input image containing a plurality of rows of pixels;
performing, by using a plurality of dedicated processing units, a pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in the input image, obtaining row data for each row of pixels; and
stitching the row data for each row of pixels, obtaining an output image, wherein
the input image contains at least one channel, and each of the at least one channel comprises a plurality of rows of pixels;
the performing a pixel extraction in parallel comprises:
padding on an edge of at least one side of the input image, obtaining a padded input image; and
performing, by using the plurality of dedicated processing units, the pixel extraction in parallel on each row of pixels of the plurality of rows of pixels in each channel of the padded input image, obtaining row processing data for each row of pixels in the channel; and
the stitching row data for each row of pixels comprises:
stitching the row processing data for each row of pixels in the channel, obtaining channel data for the channel; and
stitching the channel data for each channel, obtaining the output image.
10. The storage medium of claim 9 , wherein the computer instructions allow the computer further to implement operation of:
determining, based on a width of a convolution kernel used in performing the pixel extraction, a number of pixels being extracted each time the pixel extraction is performed on each row of pixels by each of the plurality of dedicated processing units.
11. The storage medium of claim 9 , wherein the computer instructions allow the computer further to implement operation of:
determining, based on a stride in a width direction of a convolution kernel used in performing the pixel extraction, a pixel being extracted each time the pixel extraction is performed on each row of pixels by each of the plurality of dedicated processing units.
12. The storage medium of claim 9 , wherein the computer instructions allow the computer further to implement operation of:
determining, based on a stride in a height direction of a convolution kernel used in performing the pixel extraction, an order of performing the pixel extraction on the plurality of rows of pixels by each of the plurality of dedicated processing units.