Technological Innovation World Trend
Computer vision technology includes many algorithms that can understand images, including pictures and videos. They are the foundation of many innovative key technologies-from self-driving cars to smart industrial machinery and even the software on mobile phones. They are also the capabilities we are trying to build. The foundation for machines to understand and learn the world around them as humans do themselves.
The market value of computer vision technology is expected to reach $48 billion by the end of 2022 and is likely to be the source of many ongoing innovations and breakthroughs. In a recent report, the website of Forbes magazine in the United States listed five major development and application trends of computer vision technology in 2022.
Optimize data quality
The rapid development of computer vision is thanks to the continuous advancement of deep learning technology.
Dr. Ng Enda, an important pioneer in the field of deep learning, has developed some deep learning-based image recognition models, the purpose of which is to train computers to recognize pictures of cats. These models especially rely on the quality of the data they are "fed", not just the quantity. Improving the quality of labeled data using techniques that automatically extract and label data will allow computer vision techniques to achieve the same results with less data, thereby reducing costs in terms of capital investment and computing resources, and opening up new possibilities. potential use cases.
Applied in the field of health and safety
A key application of computer vision is to spot hazards and sound an alert when something goes wrong. Scientists have developed methods that allow computers to detect unsafe behavior on construction sites, such as not wearing a hard hat, and to monitor various environments within the working range of heavy machinery such as forklifts. will automatically close. According to the data of the US Bureau of Labor Statistics, 2.7 million people are injured on the job every year, and more and more companies have increased their investment in this field to reduce the human and financial costs caused by negligence.
Of course, preventing the widespread spread of the virus is also an important use case, and computer vision technology is increasingly being used to monitor whether someone is adhering to social distancing regulations and wearing a mask. During the COVID-19 pandemic, scientists have also developed computer vision algorithms that can help diagnose patients by looking for evidence of infection and damage to images of the lungs.
for retail
In 2022, computer vision technology will be widely used in shopping and retail.
Previously, Amazon pioneered the cashierless store Amazon Go, which is equipped with cameras to easily identify items that customers take from the shelves. More stores are expected to open in 2022, joined by other retailers including Tesco, which will open its first cashierless supermarket in the UK.
In addition to automatically scanning products, computer vision has many other uses in the retail industry. For example, in the field of inventory management, cameras can check the placement of products on the shelves and the stock situation in the warehouse, and automatically order replenishment when necessary. . It is also being used to monitor and understand customer movement patterns within the store to optimize merchandise placement and, of course, to prevent merchandise theft. Another increasingly popular use case for computer vision technology is enabling customers to scan barcodes with their mobile phones to obtain product information. And in the fashion retail industry, a particularly interesting application of computer vision is the "virtual fitting room", where customers can virtually try on items without touching them, and can even identify the products that customers are trying on and provide matching suggestions .
"Show your skills" in the field of self-driving cars
Computer vision has been applied to the existing field of intelligent networked vehicles. Intelligent networked vehicles refer to equipped with advanced on-board sensors, controllers, actuators and other devices, and integrate modern communication and network technologies to realize the exchange and sharing of intelligent information between vehicles and people, roads, backgrounds, etc., and realize safe, comfortable, energy-saving and efficient driving. , and eventually a new generation of cars that can replace people to operate.
Scientists have developed some vision systems that can use cameras to track the driver's facial expressions and send out warning signals, such as the driver may be tired and may fall asleep while driving, and the survey shows that up to 25% of fatal and serious Traffic accidents are caused by this factor, therefore, such technologies and measures can better save lives.
The technology is already used in commercial vehicles such as delivery trucks, and by 2022 it is expected to find its way into private cars. Other possible uses of computer vision in the automotive field include monitoring whether occupants are wearing their seat belts or even leaving keys and phones when getting out of the car.
Of course, computer vision will also play an important role in the field of self-driving cars. As Tesla announced this year, its cars will rely primarily on computer vision, rather than using radar to model the environment around them as they drive.
Applied to the field of edge computing
Edge computing refers to the use of an open platform near the source of data to directly provide the most recent services. Edge computing is the opposite of cloud computing. Cloud computing refers to decomposing many data computing processing programs through the network, and reprocessing and analyzing these decomposed small programs through a system composed of servers to obtain results.
Edge computing technology is gaining in importance in the field of computer vision, as computer vision systems often need to make decisions quickly, such as in areas such as self-driving cars, so there is simply no time to send data to the cloud.
As computing speeds at the edge continue to increase, computer vision will have a major impact in the security arena, which is increasingly important as businesses and individuals face increased scrutiny and regulation over how video data is captured and used. Using edge devices, such as security cameras equipped with computer vision, data can be analyzed on the fly and discarded if there is no reason to keep it (such as no suspicious activity detected).