Hoh Xil Tibetan antelope detection method based on convolutional neural network

A convolutional neural network and detection method technology, applied in the field of detection of Tibetan antelopes in Hoh Xil, can solve problems such as the inability to accurately count the distribution of Tibetan antelopes, the inability to distinguish between male and female Tibetan antelopes, and the manual operation of Tibetan antelope populations. Achieve the effect of fast detection speed, good universality and good detection effect

Inactive Publication Date: 2018-09-07
XIDIAN UNIV
View PDF4 Cites 18 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, the research on the population statistics and distribution of Tibetan antelope is still in the stage of manual operation
For example, "Number and distribution of Tibetan antelope, Tibetan wild ass and wild yak in the Altun Mountain Nature Reserve" published by Lu Feiying and others from the School of Life Sciences of Capital Normal University (Journal of Beijing Normal University (Natural Science Edition), 2015 (4) :374-381.) us

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
  • Hoh Xil Tibetan antelope detection method based on convolutional neural network
  • Hoh Xil Tibetan antelope detection method based on convolutional neural network
  • Hoh Xil Tibetan antelope detection method based on convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] The present invention will be described in further detail below in conjunction with the accompanying drawings.

[0034] refer to figure 1 , the concrete implementation steps of the present invention are as follows:

[0035] Step 1, construct the training data set.

[0036] (1a) Collect a number of Hoh Xil Tibetan antelope image data under different scenes, different shooting angles, different light changes and weather conditions, and the size of the image is 1920*1080;

[0037] (1b) Perform data enhancement on the image data in (1a) above, including rotation, translation, scaling, noise perturbation and color transformation, to increase the number of training samples;

[0038] (1c) Use the image labeling tool LabelImg to label the image collected above and the Tibetan antelope image obtained by data enhancement. The labeling effect is as follows figure 2 As shown; at the same time, the coordinates, width, height and category information of the labeled Tibetan antelo...

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 present invention discloses a Hoh Xil Tibetan antelope detection method based on a convolutional neural network. The objective of the invention is mainly to solve the problems that the Tibetan antelope population statistics is slow in speed and not accurate in the prior art. The method comprises the steps of: 1) constructing a training data set, performing marking of Tibetan antelope images obtained through enhancement of the collected data, and obtaining a training data set; 2) employing a k-means algorithm to perform target frame dimension clustering of the training data set, obtaining the most effective priori frame information of the Tibetan antelope data set, and constructing a convolutional neural network with 19 convolutional layers and 5 maximum pooling layers; 3) performing training of the Tibetan antelope data by employing the convolutional neural network to obtain a Tibetan antelope detection and classification model; and 4) employing a trained model to perform detectionof the Tibetan antelope images, draw concrete positions of Tibetan antelopes and mark classes of the Tibetan antelopes. The Hoh Xil Tibetan antelope detection method based on a convolutional neural network can more rapidly and accurately perform detection and classification of the Tibetan antelopes, and can be used for scientific protection and management of the Tibetan antelopes in natural reserves.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and further relates to a method for detecting the Hoh Xil Tibetan antelope based on a convolutional neural network. It can be used for the statistics of Tibetan antelope targets in Hoh Xil Nature Reserve and the classification of male and female Tibetan antelope individuals, which is convenient for reserve management agencies to grasp the dynamics and trends of Tibetan antelopes, thereby providing a basis for scientific management of Tibetan antelope populations. Background technique [0002] Tibetan antelope, as a first-class national protected animal, is included in Appendix I of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), which is mainly distributed in alpine desert meadow grasslands and plateau grasslands in Qinghai, Tibet, Xinjiang and other places in China etc. environment. The unique ecological environment makes the protection of Ti...

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): G06K9/00G06K9/32G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V40/10G06V10/25G06N3/045G06F18/23213G06F18/214
Inventor 卢朝阳翟俊伟裴竟德李静
Owner XIDIAN UNIV
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