In a strawberry greenhouse in Belgium, we saw a very interesting picture. There was a small robot. It went through the rows of strawberries growing on the support tray. It used the machine vision detection system to find mature and intact strawberries. Then, with 3D printing claws, we gently picked each ripe strawberry and put it in the basket. If the machine vision system detects that the fruit is not yet ready to be picked, the robot estimates the time it will ripen and then returns to pick.
This machine vision inspection system can replace the traditional strawberry planting and harvesting mode, and solve the challenge of agricultural labor shortage faced by most developed countries. The robot can pick a strawberry every five seconds, while humans can pick one strawberry every three seconds.
Based on the cost constraints and other requirements for picking strawberries, the robot was designed. For example, the stem of a strawberry should not be left on the fruit when it is picked, because it will prick other strawberries in the basket. When the fruit begins to be packaged, the redder side should be placed on it to attract consumers. The robot's vision system can accomplish this task.
这台机器人的设计目的是为了与 "桌面" 生长系统配合，即草莓生长在一排排托盘上，而不是田野里，因为这是行业正在发展的方向。在欧洲，温室种植草莓已经成为一种标准方式，生产的草莓大多出口到了美国。
The robot is designed to work with the "desktop" growth system, where strawberries grow on rows of pallets, not fields, because that's where the industry is heading. In Europe, growing strawberries in greenhouses has become a standard way, and most of the strawberries produced are exported to the United States.
In addition to being easier to pick, the tray growth system is also more water-saving because it only needs to irrigate a small amount of soil around the strawberries and has a higher yield per unit area.
In the industrial field, machine vision technology has also been applied to industrial automation system to replace the traditional manual inspection to improve production quality and output. From picking and placing, object tracking to measurement, defect detection and other applications, visual data can provide simple through failure information or closed-loop control to improve the performance of the whole system.
The use of vision is not only in the field of industrial automation, we also see a large number of applications of cameras in daily life. The most basic feature of machine vision detection system is to improve the flexibility and automation of production.
In some dangerous working environment which is not suitable for manual work or artificial vision is difficult to meet the requirements, machine vision detection system is often used to replace artificial vision.
In recent years, with the rapid development of machine vision detection system, the ability of dynamic recognition and anti-interference have been greatly improved! In the future, machine vision inspection system will be more widely used in some relatively dangerous working environment and industrial production with high repeatability.