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

Tree species identification method of broad-leaved forest based on a single photo

A recognition method, broad-leaved forest technology, applied in the field of broad-leaved forest tree species recognition based on a single photo, can solve the problems that the model cannot learn high-level features, model underfitting, and few network layers

Inactive Publication Date: 2019-03-26
ZHEJIANG FORESTRY UNIVERSITY
View PDF5 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, there are still some defects. This method has too few network layers, and the model cannot learn more abstract high-level features that contain more semantic information, and this method only contains one fully connected layer, which is only suitable for processing small-scale data sets. Once the plant data increases, the model is prone to underfitting and the accuracy rate will drop

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
  • Tree species identification method of broad-leaved forest based on a single photo
  • Tree species identification method of broad-leaved forest based on a single photo
  • Tree species identification method of broad-leaved forest based on a single photo

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0073] A method for identifying broad-leaved forest tree species based on a single photo in this embodiment uses a deep convolutional neural network to learn tree species features independently, and retrains the neural network when there are enough tree species sample data sets to continuously optimize during training Neural network, after each optimization, the neural network is tested, and the convolutional neural network with the highest accuracy is selected to establish a broad-leaved forest tree species identification system, which specifically includes the following steps:

[0074] S1 collects images of different types of trees, establishes a tree species image data set, and divides the data set into a training set, a verification set, and a test set;

[0075] S2 Adjust image size: adjust each image in the tree species image dataset to an image with the same size;

[0076] S3 designs a convolutional neural network HCNN, uses the above training set images to train the con...

Embodiment 2

[0079] The broad-leaved forest tree species recognition method based on a single photo in this embodiment is based on the first embodiment. In order to ensure the accuracy of recognition, a large number of tree images need to be collected first, which can be obtained by directly shooting in a natural scene manually. It is also possible to crawl image data of related tree species in batches on the Internet by writing a crawler program. There are at least two types of trees in the tree species image dataset, and there are at least 10,000 images of each tree. Then the images in the tree species image data set are randomly divided into training set, verification set and test set, wherein the ratio of training set, verification set and test set is 5-9:0.5-2.5:0.5-2.5.

Embodiment 3

[0081] A method for identifying broad-leaved forest species based on a single photo in this embodiment, based on Embodiment 2, before inputting the image into the convolutional neural network, it is necessary to preprocess the image, specifically, the size of each image Both are modified to x×y pixels of fixed size; where x represents the width of the image, y represents the height of the image, and x=y. There is no restriction on the size of the image, just ensure that the size of each image must be consistent.

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 discloses a tree species identification method of broad-leaved forest based on a single photo, comprising the following steps: S1, collecting images of different kinds of trees, establishing tree species image data sets, and dividing the data sets into a training set, a verification set and a test set; S2, adjusting image size: adjusting each picture in the tree species image data set to an image with the same size; S3, designing a convolution neural network, training the network with the images of the training set, and testing the accuracy of the network with the images of the test set; S4 Selecting the most accurate convolution neural network model to construct a broad-leaved forest tree species recognition system, and input a tree species image for recognition, so as to obtain the recognition results. The method of the invention utilizes the depth convolution neural network to independently learn tree species characteristics, reduces artificial intervention, and has higher recognition accuracy. It can be recognized by an image of broad-leaved forest species at any angle, which is simple, flexible and practical.

Description

technical field [0001] The invention relates to a broad-leaved forest species identification method, in particular to a broad-leaved forest species identification method based on a single photo. Background technique [0002] The broad-leaved forest is the largest part of my country's economic forest. It consists of a wide variety of tree species. In addition to producing timber, its tree species can also be used to produce oil, dried fruits, industrial raw materials, medicinal materials and other by-products, etc. to a certain extent. Drive the development of China's economy. Therefore, it is necessary to deepen the understanding of broad-leaved forest species. However, in the face of such a wide variety of broad-leaved forest species, experienced people still cannot accurately identify them, let alone most people have no experience in this area. At this time, the proposal of a broad-leaved forest tree species identification method based on a single photo reduces the manual...

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): G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/2413
Inventor 冯海林胡明越杨垠晖方益明夏凯
Owner ZHEJIANG FORESTRY UNIVERSITY
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