图像分割方法、装置、计算机设备和存储介质
By combining the primary segmentation model and the secondary segmentation model, and selecting an appropriate secondary segmentation model based on the pixel ratio, the image is accurately segmented, solving the problem of inaccurate segmentation of low-resolution images in existing technologies, and achieving efficient segmentation of images of different resolutions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- INDUSTRIAL AND COMMERCIAL BANK OF CHINA
- Filing Date
- 2023-03-03
- Publication Date
- 2026-07-17
AI Technical Summary
Existing image segmentation models are not effective for segmenting images with low resolution, and may result in misidentification or omission, making them unsuitable for image segmentation scenarios with varying resolutions.
The image is initially segmented using an initial segmentation model. A secondary segmentation model is then determined based on the pixel proportions of the initial segmented regions. Different secondary segmentation models are used to perform secondary segmentation on the image, including neural network models with feature pyramid structures and deep convolutional neural network models with dilated convolutions, to achieve accurate segmentation for images of different resolutions.
It improves the accuracy of image segmentation and the generalizability of its application scenarios, making it applicable to image segmentation scenarios with different resolutions and ensuring the accuracy of the segmentation results.
Smart Images
Figure CN116228793B_ABST