Supercharge Your Innovation With Domain-Expert AI Agents!

Image processing method and image preprocessing method for target detection

An image processing and target detection technology, applied in character and pattern recognition, instruments, computer parts, etc., can solve problems such as image information loss and image information change, and achieve the effect of improving accuracy and facilitating feature extraction.

Pending Publication Date: 2020-12-29
北京豆牛网络科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the common deep learning network, it has a strong generalization ability for image quality and resolution, but it can only accept fixed-size image input in terms of image size.
The existing image size processing methods mainly have two problems: image information loss and image information change.

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
  • Image processing method and image preprocessing method for target detection
  • Image processing method and image preprocessing method for target detection
  • Image processing method and image preprocessing method for target detection

Examples

Experimental program
Comparison scheme
Effect test

example

[0079] The image processing method of the embodiment of the present invention and the image preprocessing method for training the target detection model have been specifically described above. The following description further describes the above method of the present invention in conjunction with an example. It should be pointed out that this example is only exemplary.

[0080] In this example, 300 sample images of two varieties of melon and watermelon are collected, the sizes are 480×960, 1728×2304, 325×650 and 828×1472, etc. The size of the intermediate processing image is 1000×1000, and the target detection model The size of the neural network input is 416×416. Using the methods of prior art 1, technology 2, technology 3 and the embodiment of the present invention to perform target detection through the same upgraded network. If the coincidence rate between the target detection result and the actual result is greater than 70% and the predicted category is correct, the tar...

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 an image processing method, which is used for processing a sample image into a required size, and comprises the following steps of: amplifying an image with the required size bya preset multiple in equal proportion to serve as an intermediate processing image; randomly selecting a position from the sample image in the movable range of the intermediate processing image, andprocessing according to needs so that the width and the height of the sample image are respectively not greater than the corresponding width and height of the intermediate processing image; positioning the sample image at the location on the intermediate processing image; filling the background image into all regions, except the sample image, on the intermediate processing image; and reducing thefilled intermediate processing image to a required size in equal proportion. The invention further provides an image preprocessing method for target detection using the image processing method. Imagescan be changed into the size required by target detection under the condition that the target semantic information of the sample images is not lost, and the accuracy of target detection is improved.

Description

technical field [0001] The present invention relates to the field of computer vision recognition, in particular to an image preprocessing method, device, electronic equipment and computer-readable medium for target detection. Background technique [0002] Target detection refers to the recognition and separation of objects and backgrounds in images by computers through machine learning or deep learning algorithms. Target detection is a kind of supervised machine learning. Before training the target detection model, positive and negative samples need to be prepared. Positive and negative samples play the most basic and critical role in the accuracy of target detection. In the common deep learning network, it has a strong generalization ability for image quality and resolution, but it can only accept fixed-size image input in terms of image size. There are mainly two problems in the existing image size processing methods: image information loss and image information change. ...

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): G06K9/36G06K9/00G06K9/62
CPCG06V20/46G06V10/20G06F18/214
Inventor 王勃王云吉王芳王京郑红蕾陈海英
Owner 北京豆牛网络科技有限公司
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com