Solution crystallization process image semantic segmentation method based on improved Unet model

A solution crystallization and process image technology, applied in image analysis, image enhancement, image data processing, etc., to achieve the effects of improving accuracy, reducing information loss, and reducing aliasing effects

Active Publication Date: 2021-11-16
BEIJING INSTITUTE OF PETROCHEMICAL TECHNOLOGY
View PDF14 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, inherent flaws in feature fusion prevent feature pyrami

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Solution crystallization process image semantic segmentation method based on improved Unet model
  • Solution crystallization process image semantic segmentation method based on improved Unet model
  • Solution crystallization process image semantic segmentation method based on improved Unet model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0037] A method for image semantic segmentation of solution crystallization process based on the improved Unet model, the structural block diagram is as follows figure 1 shown, including the following steps:

[0038] S1: Construct a crystallographic semantic segmentation dataset based on crystallographic image data, and preprocess the images in the dataset; preprocessing includes image cropping, correction, denoising and enhancement.

[0039] S2: The preprocessed picture is input into the improved Unet model; the improved Unet model is to add a channel-enhanced feature pyramid network between the downsampling part and the upsampling part of the Unet typical model, such as figure 2 shown.

[0040] Among them, the channel enhanced feature pyramid network includes a context enhancement module and a channel attention module:

[0041] S2.1: The structure of the context enhancement module is as follows image 3 shown, including:

[0042] Obtain the output of the first sampling ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a solution crystallization process image semantic segmentation method based on an improved Unet model. The method comprises the following steps: constructing a crystal semantic segmentation data set according to crystal image data; inputting the preprocessed picture into the improved Unet model. wherein the improved Unet model is that a channel enhanced feature pyramid network is added between a down-sampling part and an up-sampling part of a Unet typical model, according to the channel enhanced feature pyramid network, an up-down enhancement module and a channel attention module are added on the basis of a feature pyramid network; training the improved Unet model through the data set, evaluating the segmentation effect of the improved Unet model, and obtaining a trained improved Unet model; and inputting crystal image data to be segmented into the trained improved Unet model to obtain a segmentation result. Compared with a Unet semantic segmentation model based on other feature pyramid networks, the improved Unet model based on the channel enhanced pyramid can effectively improve the measurement accuracy and does not add extra complex operation.

Description

technical field [0001] The invention relates to the technical field of image semantic segmentation, in particular to a method for image semantic segmentation in a solution crystallization process based on an improved Unet model. Background technique [0002] Crystallization is the operation of separating a substance from a solution in a crystalline state. It can solve many problems that cannot be solved by operations such as unit distillation, extraction, and adsorption, and is widely used in the process of separating new products. In the crystallization process, the focus and purpose of process detection is to describe and control information such as crystal morphology and crystal size. Usually, measuring crystal morphology and size distribution during crystallization is one of the important means to control and optimize to obtain desired product quality and production efficiency. At present, process analysis tools have begun to be applied to the crystallization process, s...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06T7/10G06K9/46G06N3/04G06N3/08
CPCG06T7/10G06N3/08G06T2207/20016G06N3/045
Inventor 朱亚东洋李心超何运良李婧莹高梦楠张立立李晶崔宁
Owner BEIJING INSTITUTE OF PETROCHEMICAL TECHNOLOGY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap