Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Fruit recognition method and device based on improved SOLO network and fruit picking robot

A technology of fruit recognition and basic network, applied in the direction of picking machine, neural learning method, character and pattern recognition, etc. Strong ability to identify fast and stable effects

Active Publication Date: 2021-04-27
SHANDONG NORMAL UNIV
View PDF3 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although some progress has been made, there are still some problems in these methods: a large amount of computing and storage resources are required, and power consumption and stability issues need to be weighed; deep convolutional neural network models are generally large, the segmentation speed is not fast, and it is difficult to migrate and deploy to embedded mobile devices for real-time segmentation of green fruits under natural conditions

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
  • Fruit recognition method and device based on improved SOLO network and fruit picking robot
  • Fruit recognition method and device based on improved SOLO network and fruit picking robot
  • Fruit recognition method and device based on improved SOLO network and fruit picking robot

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0067] Embodiment 1 of the present invention provides a kind of fruit recognition method based on improved SOLO network, and this method comprises the following steps:

[0068] Obtain photos of fruit trees in an orchard environment;

[0069] Fruit tree photo is input in the identification model based on improved SOLO network, determines whether there is fruit in the described fruit tree photo;

[0070] Wherein, the recognition model based on the improved SOLO network is obtained by using multiple sets of data through machine learning training; each set of data in the multiple sets of data includes: a photo with fruit and label information identifying that the photo has fruit.

[0071] In this embodiment 1, the training based on the recognition model of the improved SOLO network includes:

[0072] Collect images of fruits in the orchard environment, and preprocess and label the images of fruits as the original data set;

[0073] Construct SOLO basic network;

[0074] Accordi...

Embodiment 2

[0098] Embodiment 2 of the present invention provides a fruit recognition device based on the improved SOLO network, the device includes: an image acquisition module, used to obtain photos of fruit trees in the orchard environment;

[0099] Recognition module, for inputting the photo of fruit tree into the recognition model based on the improved SOLO network, to determine whether there is fruit in the photo of the fruit tree; Wherein, the recognition model based on the improved SOLO network is trained by machine learning using multiple sets of data Each set of data in the multiple sets of data includes: a photo with fruit and label information identifying that the photo has fruit.

[0100] In this embodiment 2, the identification module includes:

[0101] The optimization unit is used to optimize the constructed SOLO basic network according to the identification requirements of the fruit, and replace the backbone network with the ResNeSt network to obtain an improved SOLO netw...

Embodiment 3

[0128] Embodiment 3 of the present invention provides a fruit recognition method based on the improved SOLO network, the specific process is as follows figure 1 shown.

[0129] Such as figure 1 As shown, the specific process of the fruit recognition method based on the improved SOLO network includes:

[0130] 1. Fruit image acquisition and data set production:

[0131] 1) if figure 2 As shown, the selected green fruits are green apples and immature persimmons (green). The Canon EOS 80D SLR camera is used to collect images of green fruits under different light conditions, different time periods, different weather, and different angles. The image resolution is 6000×4000 , the output format is JPG, 24-bit color image.

[0132] 2) Preprocess the collected images, reduce the image resolution to 600×400, label with LabelMe software, and construct a data set. Such as image 3 shown.

[0133] 3) The original data set is divided into training set, verification set and test set....

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 fruit recognition method and device based on an improved SOLO network, and a fruit picking robot, and belongs to the technical field of fruit picking robots. The method comprises the steps: obtaining a fruit tree picture in an orchard environment; inputting the fruit tree picture into a recognition model based on an improved SOLO network, and determining whether a fruit exists in the fruit tree picture or not; wherein the recognition model based on the improved SOLO network is obtained by using multiple groups of data through machine learning training; wherein each group of data in the plurality of groups of data comprises a picture with fruits and label information for marking the picture with the fruits. According to the fruit recognition method based on the improved SOLO network, the problem that the fruit segmentation efficiency of a visual system of a picking robot is poor is solved, the method is simple, the practice speed is high, a relatively accurate segmentation result can be obtained, and the fruit recognition method is suitable for real-time operation of the picking robot.

Description

technical field [0001] The invention relates to the technical field of fruit picking robots, in particular to a fruit identification method and device based on an improved SOLO network and a fruit picking robot. Background technique [0002] my country's fruit production ranks first in the world, but manual picking is still the main method. With the intensification of population urbanization and aging trend, the labor cost in social production is increasing day by day. The development of a fruit picking robot instead of manual picking can not only improve labor efficiency and economic benefits, but also promote the intelligent development of agricultural machinery. has strong practical significance. [0003] The rapid and accurate segmentation of target fruits directly affects the reliability and real-time performance of picking robots. The precise segmentation of objects is the key to the vision system. The picking of green fruit is a part that cannot be ignored, and it ...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08A01D46/30
CPCG06N3/08A01D46/30G06V20/10G06V20/68G06V2201/07G06N3/048G06N3/045G06F18/24G06F18/214Y02T10/40
Inventor 贾伟宽李倩雯孟虎郑元杰赵艳娜
Owner SHANDONG NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com