US 12,079,952 B2
Image processing method and system for convolutional neural network
Chia-Wei Yu, Hsinchu (TW); and Kang-Yu Liu, Hsinchu (TW)
Assigned to REALTEK SEMICONDUCTOR CORP., Hsinchu (TW)
Filed by REALTEK SEMICONDUCTOR CORP., Hsinchu (TW)
Filed on Mar. 14, 2022, as Appl. No. 17/693,723.
Claims priority of application No. 110133874 (TW), filed on Sep. 10, 2021.
Prior Publication US 2023/0081327 A1, Mar. 16, 2023
Int. Cl. G06T 3/403 (2024.01); G06V 10/774 (2022.01); G06V 10/82 (2022.01)
CPC G06T 3/403 (2013.01) [G06V 10/774 (2022.01); G06V 10/82 (2022.01)] 11 Claims
OG exemplary drawing
 
1. An image processing method for a convolutional neural network (CNN), the image processing method comprising:
obtaining input image data, a first training result, a second training result, and an interpolation lookup table, wherein the interpolation lookup table comprises a plurality of area interpolations;
segmenting the input image data into a plurality of feature blocks according to a total quantity of the area interpolations;
establishing a position mapping relationship to record the feature blocks corresponding to positions of all of the area interpolations;
assigning corresponding area interpolations to the feature blocks according to the position mapping relationship;
obtaining an interpolation parameter for each of the feature blocks according to the first training result, the second training result, and the area interpolations;
performing block convolution on each of the interpolation parameters and the corresponding feature block to obtain an output feature result; and
obtaining an output image by combining the output feature results according to the position mapping relationship.