Introduction to machine vision software and selection method

Machine vision is to use the machine instead of the human eye to make measurements and judgments. Machine vision system refers to the conversion of the ingested target into an image signal by machine vision products (ie, image capture device, divided into CMOS and CCD), and transmitted to dedicated image processing software to convert according to pixel distribution and brightness, color and other information. Digital signal; image processing software performs various operations on these signals to extract the characteristics of the target, and then controls the action of the device on the basis according to the result of the discrimination. With the vigorous development of China's manufacturing industry, China is becoming one of the most active areas of international machine vision development. The large-scale application range has gradually expanded from the beginning of electronics and pharmaceuticals to packaging, cars, etc., and has many applications in transportation and printing. The scale of use covers all occupations of the national economy. How to choose the right machine vision software has become a vital part of automation. Most of the visual enterprises can't systematically explain to the users who have just entered the line because of their shallow qualifications. As a senior company in October, we are willing to share our experience and make progress together with everyone. This article focuses on the key points of choosing machine vision software and introduces some common visual software.

机器视觉软件的种类介绍及选型方法

First, machine vision software selection points

1, the accuracy of the locator

Accurate positioning of a target or feature is an important function of a detection system or a visually guided motion system. Conventional object positioning uses grayscale values ​​to identify objects. Although this skill has been widely used, it lacks stability in the case of poor image quality. Image quality deterioration may be due to factors such as clutter, brightness, and concealment. In contrast, geometric target location is the most recent approach, which uses the generalization of the target to identify the target and its characteristics. Most of the image processing software independently developed by the VIDEO image adopts the geometric target positioning method, and the efficiency is extremely high, which can effectively help the user to solve the positioning problem.

2, tool library or application software

Machine vision software is sold in two typical ways: one is a tool library containing multiple processing algorithms, and the other is an application that specializes in a particular type of work. These two advantages and disadvantages, you need to confirm this feature in the first time before making a choice. The image processing software provided by the VIDEO image contains both types, which can provide both professional applications and diverse customer needs.

3, programming and operation is convenient

The concise and intuitive graphical interface is the key to simple application and setup. The main difference between today's machine vision products is their graphical interface. The interface is estimated from both "setup" and "operation". For an engineer, it is probably quite complicated, and it should be very simple for an operator.

4, sub-pixel accuracy

The resolution of the visual system is the smallest feature that the system can distinguish. For example, a 1'' visual scale (FOV) would give a 1/640 resolution or 0.00156'' using a 640x480 pixel computer image. In fact, machine vision algorithms have sub-pixel capabilities. That is, these algorithms can measure or derive smaller units than one pixel.

5, easy to upgrade

Machine vision systems can be used in a variety of applications, ranging from stylish cameras to surveillance systems. When choosing a system, you should consider the future upgrade of the system. The visual software system based on general purpose is very good. The end user probably thinks about the future needs of the system based on additional cameras, changes in lighting, changes in visuals, and so on.

6, image preprocessing algorithm

Most image processing software provides preprocessing algorithms, so you need to compare the richness, efficiency, etc. that they provide.

7, system integration

If you are not very proficient in machine vision skills, then a system integrator is needed for your project, and the ideal visual product can be widely accepted by system integrators.

Second, the introduction of common machine vision software

1. OpenCV, open source free image processing library

OpenCV is an open source, free computer vision library launched in recent years. It can easily implement digital image and video processing with its included functions. At the same time, using the object-oriented VC++ 6.0 programming tool, programming in C++ language OpenCV algorithm library provides great convenience for VC++ programming digital image processing, which will become a powerful tool in the field of image and video processing.

Disadvantages: Because it is open source software, its version is numerous, the function library is complex, and the execution efficiency is applied. It is more suitable for scientific research and learning, and is not suitable for industrial applications.

2. Halcon, a powerful image processing library

HALCON is a complete set of standard machine vision algorithm packages developed by MVtec, Germany, with a wide range of machine vision integrated development environments. It saves product costs and shortens the software development cycle – HALCON's flexible architecture facilitates rapid development of machine vision, medical image and image analysis applications. The industry of Europe and Japan is already recognized as the best performing Machine Vision software.

Disadvantages: The price is more expensive, and each time you need to re-purchase the license.

3. NI Vision, a fast-verified image processing library (including visual assistant, VBAI)

Electronic Wire

Electronic Wire


Feyvan electronics wires and cables are used to industrial and automotive field including home appliance, sensor, motor, rotary encoder, automotive wiring assemblies etc. For the purpose of providing excellent service, our engineers and design specialists devote to help you choose and design standard or custom wire and cable to your unique application.


The cables are available in a wide variety of configurations and features. Conductors include annealed stranded tinned copper wire or bare copper. Insulations include PVC, Silicone Rubber, XLPE, Teflon and etc., all of which are RoHS, REACH, UL and CSA compliant. We have complete production lines to satisfy customers' needs.

Electronic Wire,Electronic Wiring,Electrical Cable,Electrical Wiring

Feyvan Electronics Technology Co., Ltd. , https://www.fv-cable-assembly.com