US 12,094,193 B2
Image processing method and image processing device
Yingying Li, Beijing (CN); and Zhenglong Li, Beijing (CN)
Assigned to BOE Technology Group Co., Ltd., Beijing (CN)
Filed by BOE Technology Group Co., Ltd., Beijing (CN)
Filed on Sep. 1, 2021, as Appl. No. 17/463,898.
Application 17/463,898 is a continuation of application No. 15/974,926, filed on May 9, 2018, granted, now 11,138,738.
Claims priority of application No. 201711052903.4 (CN), filed on Oct. 31, 2017.
Prior Publication US 2021/0398287 A1, Dec. 23, 2021
Int. Cl. G06V 10/82 (2022.01); G06F 18/2413 (2023.01); G06N 3/02 (2006.01); G06T 7/00 (2017.01); G06T 7/11 (2017.01); G06V 10/40 (2022.01); G06V 10/764 (2022.01); G06V 10/44 (2022.01)
CPC G06V 10/82 (2022.01) [G06F 18/2413 (2023.01); G06N 3/02 (2013.01); G06T 7/0002 (2013.01); G06T 7/11 (2017.01); G06V 10/40 (2022.01); G06V 10/764 (2022.01); G06F 2218/08 (2023.01); G06T 2207/20076 (2013.01); G06T 2207/20081 (2013.01); G06V 10/454 (2022.01); G06V 2201/03 (2022.01)] 14 Claims
OG exemplary drawing
 
1. A method of processing a dermoscope image, comprising:
transforming a first dermoscope image to obtain two or more second dermoscope images for the first dermoscope image;
performing feature extraction on the two or more second dermoscope images by using a first machine learning unit selected from a group of first machine learning units, and obtaining a feature map of each of the two or more second dermoscope images; and
inputting the feature map of each of the two or more second dermoscope images to a second machine learning unit to obtain a processing result of the first dermoscope image,
wherein the obtaining a feature map of each of the two or more second dermoscope images comprises: outputting feature maps having a same resolution of each of the two or more second dermoscope images,
wherein the first machine learning unit comprises a first neural network, and the method further comprises:
outputting the feature maps having the same resolution of each of the two or more second dermoscope images through at least one intermediate layer of the first neural network of the selected first machine learning unit,
wherein the first neural network comprises an Inception v3 neural network or an Inception-ResNet-v2 neural network,
wherein the outputting the feature maps having the same resolution of each of the two or more second dermoscope images through at least one intermediate layer of the first neural network of the selected first machine learning unit comprises:
outputting the feature maps having the same resolution of each of the two or more second dermoscope images through at least one intermediate layer of the Inception v3 neural network or the Inception-ResNet-v2 neural network of the selected first machine learning unit.