6 Parameters that Influence Speed of Machine Vision Inspection

By March 11, 2022November 28th, 2023Image Acquisition, Image Processing
Speed of machine vision

Cycle time in terms of industrial processes refers to the time in which a unit of any task is completed so that the flow system can realize the target quantity output within a certain period. The cycle time is calculated by the daily net-working time divided by the number of units of work performed in the day. The cycle time of a machine vision system is also calculated similarly. The demands of the industry have greatly reduced cycle time with the help of automation and the integration of new technology leading to growing demand for high speed Machine Vision Inspection.

The speed of machine vision inspection is largely dependent on – 

Material Handling Time – 

The material handling time is the time between which a material to be inspected is exposed in front of the image acquisition medium in a way that it can focus adequately on the material for acquiring images. In an industrial setting, the materials are usually on an assembly line or conveyor belt. The camera is either fixed or movable and is placed at a certain point of the assembly line. When the material enters the field of focus of the camera, the material handling time begins and ends when the material is completely in focus. This usually takes a few milliseconds. 

Trigger Delay – 

Trigger Delay is the time between which the trigger has been sent to the image acquisition system to the time at which the actual image acquisition process begins. In the conveyor belt scenario, the trigger delay is the time between the object that passed through the proximity sensor to the time in which the image of the object is captured. This delay is in the range of microseconds to a few milliseconds.

Also read: Image Acquisiton Trigger Mechanism

Image Capture Time – 

The image capture time is the time in which the actual image acquisition happens. The time taken by the sensors of the camera to first begin capturing the data from the object to the time in which all possible information is extracted from the object comes under image capture time. The light information from the sensors is converted to electrical information during this time interval which is then digitally transformed into an image in the future steps. This is also called the image acquisition time or image exposure time as this is the time in which the sensors are exposed to the object. 

Image Transfer Time – 

The image transfer time refers to the time between which the image gets transferred from an image acquisition system to an image processing system via an interface. Usually, the camera transfers the image to a computing system via ethernet or USB cables. In the case of smart camera based vision systems where the processing system is in-built into the camera, this component of time is minimal and can be ignored. But in most cases, a host computer receives the image formed in the camera through a wired medium and this time is considered in the cycle time of the machine vision model. 

Image Processing Time – 

The image processing time is the time taken by the computer or an image processing system to process the image via algorithms that are rules-based or AI-based. Multiple software operations take place during this time and they provide a report or inference of the image. Hence this time duration is also called image inferencing time. This process is the important step to speed up machine vision inspection and complex processes happening inside the computer to gain the result of the image processing that has happened.

Result Communication Time – 

The result communication time is the time difference between image processing result generation to the reception of this result by a control system through some interface. As this is a cross-medium time, the speed and efficiency of this process are determined by the interface through which the communication takes place. For instance, when a host computer sends a signal to the control system via a serial port, the latency of the serial port needs to be considered into the communication time delay. After this communication is done, the control system stores the result or performs some action based on the result. 

Summary –

The cycle time can have bottlenecks or transition losses and it is important to detect where they are coming from and how to fix them. For instance, image acquisition time is not very high, but image transfer time can be high because it involves the movement of data from one physical location to another. The image processing also occupies a large portion of the cycle time because of the crucial actions performed in it. 

Thus, the speed of machine vision inspection is determined by the cycle time. The cycle time is unevenly distributed among 6 components that have varying speeds and transition processes. The best possible way to optimize the cycle time of the machine vision system is to detect bottlenecks and transition losses and minimize them. When this becomes possible, the end-to-end time of the machine vision system reduces significantly and stays within the limits of the targeted time required to perform a certain action.

2 Comments

Leave a Reply

Schedule A Demo
close slider


hbspt.forms.create({ region: "na1", portalId: "726123", formId: "bd9f048a-7e04-4512-9179-c50855c961ea" });