The real AI Taiho color sorter that can learn is now 1 year old.
Release Date：2020-04-17 Source： Number of views：533
MachineLearningMachineLearning
You must still remember the world battle between AIphaGo and Go world champion Lee Sedol in 2016, which caused a stir when the artificial intelligence AIphaGo finally won 4:1.
In fact, as an AI robot that beat the world champion, its main working principle is "deep learning".
With the development of computer technology, AI technology has been more widely used to create more value for all walks of life.
Taiho Optoelectronics & AI Technology
Taiho Optoelectronics has applied AI technology to a food color sorting project in Europe as early as 2019. This project utilizes machine deep learning to give the AI program the ability to recognize images, which fills the gaps in the field of AI sorting at home and abroad; it satisfies the accurate sorting of large materials that cannot be met by ordinary color sorting technology, brings more color sorting convenience to customers and promotes the increase of production.
(Taihe AI color sorter)
Deep learning simply refers to the deep neural network that simulates the human brain for analysis and learning, i.e., the computer program is used to simulate the human learning ability to acquire knowledge and experience from real examples.
Taiho Optoelectronics realizes digital image processing through convolutional technology and uses neural network to learn the abstract features of the image to provide deep learning detection, segmentation and recognition algorithms for material identification, and realizes the precise rejection of materials with industrial-grade high-speed computing platform.
Taihe AI Color Sorter
Taiho's application of AI technology to the field of color sorting is also a milestone innovation after both Taiho's first application of CCD technology to the color sorting industry.
Taihe AI color sorter is realized through the following four steps.
First of all, the machine labels the categories of these material images through a large number of material sample images collected;
Second, the samples are labeled well for computer algorithm learning to get the standard material model;
Subsequently, the model is used to detect, segment, and identify the new material image, such as defects, shape, and color of the material;
Finally, Taihe AI color sorter, eliminates the bad materials and sorts out the materials that meet the standard.
(Neural network layer structure)
Taiho Photonics AI Application
In Taiho photoelectric AI color sorting equipment, in the computer vision technology and deep learning technology used in material sorting scenarios with high recognition accuracy, classification recognition rate of 99.99% or more.
To sum up, Taiho photoelectric AI color sorting equipment adopts high-precision deep convolutional network image recognition algorithm, which has the features of high recognition accuracy and high-speed algorithm, and is able to control the recognition time of a single material picture within 5ms (i.e., more than 720,000 material pictures can be processed per hour), which can effectively guarantee the processing capacity of the equipment.
In the future, Taiho Optoelectronics will continue to innovate technology, adhere to innovation, lead the development of the color separation industry, and help enterprise customers to drive industrial upgrading by technology.
MachineLearningMachineLearning
You must still remember the world battle between AIphaGo and Go world champion Lee Sedol in 2016, which caused a stir when the artificial intelligence AIphaGo finally won 4:1.
In fact, as an AI robot that beat the world champion, its main working principle is "deep learning".
With the development of computer technology, AI technology has been more widely used to create more value for all walks of life.
Taiho Optoelectronics & AI Technology
Taiho Optoelectronics has applied AI technology to a food color sorting project in Europe as early as 2019. This project utilizes machine deep learning to give the AI program the ability to recognize images, which fills the gaps in the field of AI sorting at home and abroad; it satisfies the accurate sorting of large materials that cannot be met by ordinary color sorting technology, brings more color sorting convenience to customers and promotes the increase of production.
(Taihe AI color sorter)
Deep learning simply refers to the deep neural network that simulates the human brain for analysis and learning, i.e., the computer program is used to simulate the human learning ability to acquire knowledge and experience from real examples.
Taiho Optoelectronics realizes digital image processing through convolutional technology and uses neural network to learn the abstract features of the image to provide deep learning detection, segmentation and recognition algorithms for material identification, and realizes the precise rejection of materials with industrial-grade high-speed computing platform.
Taihe AI Color Sorter
Taiho's application of AI technology to the field of color sorting is also a milestone innovation after both Taiho's first application of CCD technology to the color sorting industry.
Taihe AI color sorter is realized through the following four steps.
First of all, the machine labels the categories of these material images through a large number of material sample images collected;
Second, the samples are labeled well for computer algorithm learning to get the standard material model;
Subsequently, the model is used to detect, segment, and identify the new material image, such as defects, shape, and color of the material;
Finally, Taihe AI color sorter, eliminates the bad materials and sorts out the materials that meet the standard.
(Neural network layer structure)
Taiho Photonics AI Application
In Taiho photoelectric AI color sorting equipment, in the computer vision technology and deep learning technology used in material sorting scenarios with high recognition accuracy, classification recognition rate of 99.99% or more.
To sum up, Taiho photoelectric AI color sorting equipment adopts high-precision deep convolutional network image recognition algorithm, which has the features of high recognition accuracy and high-speed algorithm, and is able to control the recognition time of a single material picture within 5ms (i.e., more than 720,000 material pictures can be processed per hour), which can effectively guarantee the processing capacity of the equipment.
In the future, Taiho Optoelectronics will continue to innovate technology, adhere to innovation, lead the development of the color separation industry, and help enterprise customers to drive industrial upgrading by technology.
