Large volume parenteral foreign body detection method, system, medium and equipment based on deep learning and target tracking

A target tracking and deep learning technology, applied in the field of medical image detection, which can solve the problems of high missed detection rate, high maintenance cost, and visual fatigue.

Pending Publication Date: 2020-11-03
湖南爱米家智能科技有限公司
View PDF0 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The artificial light inspection method used in traditional small and medium-sized pharmaceutical companies relies on manpower and has the disadvantages of inconsistent detection standards, resulting in visual fatigue, slow detection speed, and hi

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
  • Large volume parenteral foreign body detection method, system, medium and equipment based on deep learning and target tracking
  • Large volume parenteral foreign body detection method, system, medium and equipment based on deep learning and target tracking
  • Large volume parenteral foreign body detection method, system, medium and equipment based on deep learning and target tracking

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0062] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0063] In this example, the camera adopts an area array gigabit network CCD camera (Baumer TXG12) with a resolution of 1080*960, the lens is a 6mm wide viewing angle Computar lens, and the light source is a dome diffuse reflection light source with a radius of 6cm (LTS-FM12030-WQ) ;

[0064] Such as figure 1 As shown, a large infusion foreign body detection method based on neural network and target tracking includes the following steps:

[0065] Step 1: Collect the sequence images after high-speed rotation-emergency stop on the large infusion production line: denoted as Image0-7.

[0066] Step 2: After performing format conversion on the large infusion sequence images collected in step 1, cut out the boundary interference area: denote as the preprocessed image ImageI_0-7.

[0067] Step 2.1: First convert the historical sequence images of large infusi...

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 large volume parenteral foreign body detection method, a system, a medium and a device based on deep learning and target tracking, and the method comprises the steps: carrying out the image preprocessing of a plurality of collected continuous images, combining a target detection algorithm and a target tracking algorithm, and achieving the fusion of target detection and target tracking and the precise positioning and tracking of foreign bodies; firstly, preproce3ssing a sequence image, then using a Faster RCNN neural network for carrying out target detection on a firstframe of image to obtain the initial position of each suspected target, then tracking the positions of each target in later frames through a CSRDCF target tracking algorithm to obtain the motion trail of each suspected target, and obtaining the motion trail of each suspected target, and finally, according to the track characteristics, performing classification by using an adaptive classificationalgorithm based on a semi-naive Bayes principle, and eliminating noise interference. Experiments show that the method not only can greatly improve the detection speed, but also can greatly improve thedetection precision, and meets the requirements of industrial production precision and real-time performance.

Description

technical field [0001] The invention belongs to the field of medical image detection, and relates to a method, system, medium and equipment for detecting foreign objects in large infusion fluids based on deep learning and target tracking. Background technique [0002] my country's infusion solution production and sales volume has already ranked first in the world. In the pharmaceutical production process, due to various reasons, tiny foreign objects with a diameter greater than 50 microns are mixed into the drug solution, including fibers in the air and hair dropped by workers. , glass shavings produced by the collision of glass bottles, rubber blocks from rubber stoppers, etc. Because the large infusion solution directly injects the drug into the vein and enters the blood of the human body, once these foreign objects enter the human blood circulation, it is directly related to the safety of human life. The resistance is much weaker. Once the medicine with foreign matter is ...

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): G06T7/246G06T7/44G06T7/90G06T7/10G06T7/00G06K9/62
CPCG06T7/246G06T7/44G06T7/90G06T7/10G06T7/0012G06T2207/10004G06T2207/20132G06T2207/20081G06T2207/20084G06T2207/30104G06F18/2415
Inventor 张辉王群易俊飞毛建旭周显恩朱青王耀南
Owner 湖南爱米家智能科技有限公司
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