How to choose machine vision software?

With the development of the automation industry, there are many types of machine vision software industries. How to choose a machine vision software that suits your needs? Below, the editor of Chengdu Xinxiwang will explain to you 10 must-know elements.

1. Locator     

The precise localization of objects or features is an important function of a detection system or a vision-guided motion system. Traditional object positioning uses gray value correlation to identify objects. Although this technology has been widely used, it lacks stability when image quality deteriorates. Deterioration in image quality may be due to factors such as clutter, brightness differences, and occlusion. In contrast, geometric object localization is a new method that uses the object's outline to identify the object and its features.

2. Optical devices and lighting

It is well known that proper optics and lighting are critical to the success of vision applications. Sometimes, despite the selection of appropriate optics and lighting, slight changes in the object or feature being monitored require corresponding changes in lighting and lighting intensity. For example, there are differences in the polished surface of the wafer and the quality of laser-etched marks in OCR applications. A stable positioning tool can easily handle situations where image quality deteriorates due to variations in light concentration and brightness. Using the appropriate software allows you to eliminate the need for dimming operations regardless of how the image quality changes.

3. Complete tool set for multiple tools tied together

Machine vision software is mainly sold in two typical forms: a complete vision toolset, and an application of tools for a specific task, such as BGA inspection. The user's application will determine whether to use a complete visual toolset or a large number of specific tools. A vision tool is a general application or algorithm that can perform a predetermined task on an image or a certain part of an image. For example, a blob detection tool can find a group of dark or light pixels and measure the blob's various dimensions. The key to choosing a vision system is to have a complete set of vision tools. Although you won't need all the tools at the beginning of your project, after a few years your requirements will change and you may need additional tools for new applications.

4. Easy to program and operate

A simple, intuitive graphical interface is key to ease of use and setup. The main difference between today's machine vision products is their graphical interface. The interface should be evaluated from the two aspects of "settings" and "operation". It should be very complex for an engineer and very simple for an operator.

5. Sub-pixel accuracy

The resolution of a vision system is the smallest feature that the system can resolve. For example, "A field of view (FOV) of 1 using a 640x480 pixel computer image will result in a resolution of 1/640 or 0.00156". In fact, machine vision algorithms have sub-pixel capabilities. That is, these algorithms are able to measure or derive units smaller than a pixel.

6. Future upgrades

Machine vision systems are used in a variety of applications, ranging from stylish cameras to surveillance systems. Future upgrades to the system should be considered when selecting a system. Universal vision software enables users to configure the appropriate lighting, optical systems and vision tools. The machine vision image processing software independently developed by Weishi Image has a wide range of applications. Its MVIPS machine vision image processing software provides multiple library functions such as image measurement, character recognition, color analysis, defect detection and target positioning. Users can use it to quickly develop images. Process and build your own machine vision application system.

General-purpose vision software systems are better upgraded. Users should consider future needs for the system in terms of additional cameras, changes in lighting, changes in vision tools, etc.

7. Image preprocessing

It is very important to detect feature points and defects regardless of brightness and differences in object surface or material. Image preprocessing algorithms can enlarge the feature points of an image so that visual tools can better detect them. Likewise, feature points can be reduced so that vision tools ignore them.

8. Visually guided movement

If your application requires a vision system to guide the robot, you must know how the vision system is integrated with the motion system. For calibration and operation, the unintegrated motion system and vision system are preliminary systems, and the robot or mechanism and vision system are calibrated separately. In operation, an independent vision system calculates an offset from the part's position based on a known position in the vision coordinate system, and then instructs the robot's arm to pick up the part at an offset from the initial programmed pick position. .

9. System integration

If you are not very proficient in machine vision technology, then you need a system integrator for your project. The ideal vision product can be widely accepted by the system integrator.

10.Factory level connection

Currently, there are various methods of communicating with vision systems, common interfaces, such as serial port (RS-232), RS-485, parallel port, Ethernet, Devicenet, digital I/O, etc. Newer interfaces such as IEEE-1394 and USB are also widely used. When evaluating vision systems, consider factory floor connectivity. Typically, a machine vision system is a data acquisition system that interfaces with other factory floor equipment and the factory's information system. Some vendors' software enables remote operation of vision systems over a local area network or the Internet. In certain industries, such as pharmaceuticals, machine vision systems are required to be self-contained for specific applications, ensuring that the setup is not disrupted by remote operations. When selecting a vision system, the vision system's communication interface is an important consideration and should not be overlooked.