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

Target detection method and device for shellfish images

An image detection and target detection technology, applied in the field of deep learning, can solve the problems of inability to deploy, high hardware computing power requirements, and low accuracy, and achieve the effect of providing technical feasibility, high detection efficiency and accuracy

Inactive Publication Date: 2019-09-20
CHINA AGRI UNIV
View PDF2 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, target detection algorithms based on deep learning methods have achieved certain results in academia and industry, but there is still little research in the field of agriculture. In the existing target detection methods, due to the complex network structure, the The hardware computing capability is high, and it cannot be deployed on the mobile terminal. There are problems such as low detection efficiency and low accuracy, and it is difficult to meet market demand.

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
  • Target detection method and device for shellfish images
  • Target detection method and device for shellfish images
  • Target detection method and device for shellfish images

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0020] Among the existing target detection algorithms based on deep learning methods, three target detection models, Faster R-CNN, YOLO v3 and SSD, are commonly used. Among them, Faster R-CNN aims at the time-consuming problem of candidate frame extraction in target detection. In the step of extracting candidate frames, a neural network for edge e...

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 embodiment of the invention provides a target detection method and device for shellfish images. The target detection method comprises the following steps: acquiring a to-be-detected target shellfish image; detecting the target shellfish image based on the trained shellfish image detection model to obtain a target detection result of the target shellfish image; , wherein the trained shellfish image detection model is constructed by replacing a VGG16 network in an SSD framework with a MobileNet network and is obtained by training sample shellfish image marked with shellfish species; the shellfish local feature extraction method based on the MobileNet network is characterized in that the shellfish local feature extraction method based on the MobileNet network is adopted, and is obtained by training shellfish images marked with shellfish types, and a residual attention mechanism is arranged between the MobileNet network and the prediction network, so that the local feature acquisition capability is improved. According to the embodiment of the invention, an SSD target detection basic framework is adopted; a shellfish image detection model is constructed by combining a MobileNet lightweight network, so that shellfish images are detected, shellfish types in the shellfish images can be quickly identified, compared with an existing shellfish detection method, the shellfish image detection method has higher detection efficiency and accuracy, and technical feasibility is provided for shellfish image detection deployed by a mobile terminal.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to a target detection method and device for shellfish images. Background technique [0002] Aquatic products represented by marine shellfish are rich in protein and vitamins, and have the characteristics of low fat and good nutritional balance. They have become the main source of people's intake of high-quality animal protein. The vast consumer market has promoted the healthy, stable and sustainable development of my country's shellfish farming industry, and played an important role in optimizing the dietary structure of residents, promoting economic development and increasing the income level of practitioners. [0003] As an important part of the agricultural economy, marine shellfish has the characteristics of various types and complex characteristics. The traditional human-based detection and recognition operations for shellfish are mostly based on manual extraction of cont...

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
CPCG06N3/08G06N3/048G06N3/045G06F18/241
Inventor 李振波李光耀张楚悦李耀东岳峻李道亮
Owner CHINA AGRI 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