Lactating sow posture recognition method based on improved Faster-R-CNN

A technology for sucking sows and recognition methods, applied in the field of target detection and recognition, to overcome the influence of scene light changes, improve gesture recognition performance, and increase time costs

Active Publication Date: 2018-11-16
SOUTH CHINA AGRI UNIV
View PDF8 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, there are few reports in the literature on the use of co

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
  • Lactating sow posture recognition method based on improved Faster-R-CNN
  • Lactating sow posture recognition method based on improved Faster-R-CNN
  • Lactating sow posture recognition method based on improved Faster-R-CNN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0051] figure 1 The first part is the establishment of the depth image database, including RGB-D video image acquisition, depth image preprocessing, data set annotation to obtain the original training set and test set, and the preparation of the training set for the expansion of the original training set, and the final labeled training set and The test set constitutes a deep image database to provide data support for subsequent model training and testing. The second part is to design a robust, real-time and high-precision CNN structure. Firstly, the ZF network with strong real-time performance is selected as the basic structure, and then the network depth is increased and the residual structure is introduced to complete the structural design. The third part is to design an improved Faster-R-CNN sow gesture recognition model. By using the convolutional layer of the CNN structure designed in the second part as the shared convolutional layer of the Faster-R-CNN network, its fully...

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 relates to a lactating sow posture recognition method based on an improved Faster-R-CNN. The method comprises the following steps: S1, an RGB-D video image of a lactating sow is acquired, and a lactating sow posture recognition depth video image library is built; S2, the depth of a basic ZF network is added, a residual structure is introduced, and a CNN network structure with high accuracy, real-time performance and robustness is designed; S3, the designed CNN network structure is used to build a Faster-R-CNN model structure, Center Loss supervisory signals are introduced to theFaster-R-CNN model structure, in joint with SoftmaxLoss, a classification loss function is built, and finally, an improved Faster-R-CNN lactating sow posture recognition model is built; and S4, a training set is used to train the Faster-R-CNN lactating sow posture recognition model, a test set is used to test the model performance, and a model with the best performance is finally selected for lactating sow posture recognition.

Description

technical field [0001] The present invention relates to the field of target detection and recognition in computer vision, and more specifically, relates to an improved design of CNN network structure, based on the Faster-R-CNN target detection algorithm, and a lactating sow gesture recognition method that introduces Center Loss supervision signals. Background technique [0002] Maternal behavior and health and welfare of sows directly affect the economic efficiency of pig farms. Automatic recognition of sow posture is an important basis for early warning of high-risk movements of sows, automatic analysis of sows' nesting behavior, automatic monitoring of nursing piglets, and assessment of health and welfare status. Using computer vision to automatically monitor pigs is not only low-cost, high-efficiency, and non-destructive, but also avoids the stress response of pigs caused by sensor monitoring methods, and is gradually applied to pig posture recognition such as standing, 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): G06K9/00G06K9/62
CPCG06V40/10G06F18/285G06F18/214G06F18/24
Inventor 薛月菊朱勋沐郑婵陈鹏飞杨晓帆
Owner SOUTH CHINA AGRI 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

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