What is machine vision inspection?

Quality Inspection Using Machine Vision

What is machine vision inspection?

An inspection machine vision system is a system that uses cameras and computers to inspect manufactured products for any type of nonconformance during the manufacturing process. An example might be inspecting a piece of silicon wafer for processed induced defeats post photolithography process, such as uneven dispense or exposure issue with the stepper. There are two different inspection methods. The inspection requires you to study the picture frame or product through the camera of the machine vision system. Depending on the amount of variation in the picture you are inspecting, you can either: Retrace and search the path in the camera, or Create a statistical sample from which to identify the commonalities Automatic scanning One type of machine vision system, called scanning, is designed for automatically scanning the product frame and scene. The scanning frame also varies depending on the scanning solution. For example, bar code scanners require a product to move by certain percentage, a lighted tag reader can use optical lenses, and other optical readers require exacting alignment of the product and the reader.

How does machine vision inspection work?

Many physical inspection techniques rely on an inspection camera that takes photographs or videos of the part or process being inspected. These images are then turned into measurements that the technician can assess visually to see if the parts meet requirements or are affected by problems. For example, a laser level is used to capture images of the surface of a product being evaluated. This image data is then put through a computer to produce measurements of the height, depth, and/or angle of the parts. This measurement data is then used by the technician to compare the measurements to the specifications for quality or inspect for presence of defects or functional flaws.

Why is machine vision inspection necessary?

Product defects need to be detected and eradicated before they damage the reputation of a manufacturer. Product defects need to be detected and eradicated before they damage the reputation of a manufacturer. In such cases, there is a great deal of impact on revenue, whether from civil lawsuits, product recalls, recalls by distributors, or product recalls by end users. For example, last year, Microsoft had to recall several laptops with a faulty webcam that could have been hacked to spy on users, causing an estimated $4.3 million in costs. Manufacturers also need to use machine vision inspection to meet the requirements of regulations for food and pharmaceutical products.

Common machine vision inspection tasks?

Machine vision inspection tasks may include:

  • Particle Analysis: Checking for particulate matter or contaminants on parts
  • Read/Mark: Identifying and reading barcodes and other markings on products
  • Examine/Measure: Measuring physical dimensions of an object
  • Screening: Checking products against predefined criteria or requirements
  • Surface inspection, include surface damages, scratches, and contamination
  • Robot task inspection, such as the quantity and quality of glue beads for the automotive industry  
  • Object Positioning, such as the placement of Barcode and label.
  • Inspection of medicine vials for level, color, and mixture ratio.


Machine vision inspection is the detection of defects on objects, such as semiconductor wafers or printed circuit boards, by analyzing digital images. It is a type of non-contact optical sensor that uses a digital camera and computer to inspect parts. Machine vision inspection technology can detect flaws such as scratches, particles debris and ink marks. Machine Vision has really changed manufacturing over the past 20 years, and will continue to help factories improving quality, productivity, and throughput while driving down manufacturing costs.

Need Help with your machine vision project today?


Leave a Comment

Your email address will not be published. Required fields are marked *