EasyMv visual detection system is an embedded visual detection system developed by Beijing Mwei Technology Studio. The purpose of this system is to simplify the visual detection process, reduce the difficulty of visual detection, lower the cost of visual detection, and promote the further popularization of machine vision in industrial production.
EasyMv visual detection system consists of visual detection software and development board. All algorithms in the visual detection software are independently developed by the studio, with independent intellectual property rights. It adopts a C++ combined with QT development framework, which has the advantages of high detection running efficiency and low resource consumption, facilitating the extension of the visual detection architecture to systems such as Windows and Android. The development board uses the mainstream open-source development board Raspberry Pi 4B (4G), which has the advantages of good development ecology and stable operation. This development board supports various modes of IO interfaces, capable of meeting the triggering detection and result output functions for 4 cameras.
The detection project has been under development since 2020, with a cumulative shipment of 3000 units, covering size detection, presence detection, and other projects.
In 2024, Mwei Technology Studio will increase investment in the EasyMv project, with the following major improvements:
1. Based on the original support for conventional industrial cameras such as Hikvision, Meidwei Vision, and Dushen, we have added driver-free cameras, further reducing system costs. For example, this time we used a driver-free camera purchased for 13 yuan on a certain treasure, which can complete the construction of a visual project.
2. Added a dedicated IO interface board, making it easier to complete IO wiring.
The overall system assembly diagram is shown in the figure.
Today, this system was used to complete a detection project for presence detection, with the main process as follows:
-
Open the software
-
Create a new detection box and set the Blob analysis method for area detection
-
Set the area reference value, if greater than the set range, it is considered that the pattern exists, outputting IO0; otherwise, outputting IO1 for the reverse side.
The visual scheme construction process is shown in the video:
Let’s review the previous project introduction articles of EasyMv.
Detailed introduction to the EasyMv visual detection system (based on Raspberry Pi)
Implementation of image calibration algorithm based on Raspberry Pi template positioning
Raspberry Pi for jumper cap defect detection
Raspberry Pi + QT for two-dimensional size measurement
Template matching of the visual detection system based on Raspberry Pi
If you have project needs, feel free to scan the QR code for Mwei Technology Studio.
