US 12,380,567 B2
Image processing
Yuying Hao, Beijing (CN); Yi Liu, Beijing (CN); Zewu Wu, Beijing (CN); Baohua Lai, Beijing (CN); Zeyu Chen, Beijing (CN); Dianhai Yu, Beijing (CN); Yanjun Ma, Beijing (CN); Zhiliang Yu, Beijing (CN); and Xueying Lv, Beijing (CN)
Assigned to BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD., Beijing (CN)
Filed by BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD., Beijing (CN)
Filed on Nov. 23, 2022, as Appl. No. 18/058,543.
Claims priority of application No. 202111424250.4 (CN), filed on Nov. 26, 2021.
Prior Publication US 2023/0085732 A1, Mar. 23, 2023
Int. Cl. G06T 7/11 (2017.01)
CPC G06T 7/11 (2017.01) [G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An image processing method, the method comprising:
obtaining an image to be processed that includes a target region to be annotated;
in response to a first click on the target region, performing a first operation, wherein the first operation expands a predicted region for the target region based on a click position of the first click, and reduces an area of the target region that has not been covered by the predicted region;
in response to a second click on a position where the predicted region exceeds the target region, performing a second operation, wherein the second operation reduces the predicted region based on a click position of the first click a click position of the second click, and reduces an area of the predicted region exceeding the target region; and
in response to determining that a difference between the predicted region and the target region is less than a threshold, obtaining an outline of the predicted region to annotate the target region,
wherein the first operation includes:
inputting the click position of the first click and the image to an image processing model, wherein the image processing model is trained with sample data including a sample image and a target region in the sample image; and
obtaining an expanded predicted region output by the image processing model, and
wherein the second operation includes:
inputting the click position of the first click, the click position of the second click and the image to the image processing model; and
obtaining a reduced predicted region output by the image processing model.