globe-alt Visit our US site Contact
EN
DE
ES
FR
PT
user Login

Smart cameras for machine vision

Smart cameras – also known as intelligent cameras – combine a sensor, processor and evaluation software in compact housing. They deliver results directly to your control system and are ideal for clearly defined tasks such as presence checks, code reading, or quality controls.
 
STEMMER IMAGING can help you select the right smart camera solutions for your application. This page provides an overview of the available models and helpful information to assist you in making your decision. 

Explore our Smart Camera Models

  • ALECS_front_left_view_sRGB
    Allied Vision - Alecs
    Image Processing with Alecs: Power, Precision, and Performance in One Solution!
  • Gocator2DCamera_isoR_comp_R
    LMI - Gocator 2D
    High-speed measurement and inspection – compact, PC-free, and ready for industrial deployment.

2 Items

Not the right smart camera here?

If the right smart camera for your application is not listed here, please contact us. Together we will find the right solution for your task.

More than just a smart camera. Your project, our service

 

At STEMMER IMAGING, you get more than hardware. We help you with your selection, supply coordinated components, and support you with integration, configuration, and optimisation. This means you get a solution that is exactly tailored to your project.

Questions about our smart cameras? We are happy to advise you

 

Our specialists will be happy to help you select the right smart camera – from resolution and interfaces to software integration.

 

Get in touch with us and we will guide you step by step.

Choosing the right smart camera

 

Before you decide on a model, there is always the fundamental question: Is a smart camera the right choice for your application? For clearly defined tasks – such as reading codes, presence checks or simple quality controls – this is usually the case. However, if several cameras need to work synchronously, very high resolutions or complex evaluations are required, smart cameras reach their limits, and a complete vision system is the better solution.

 

The following guide will help you select the appropriate criteria in the product filter and thus find the right camera step by step.

Step 1: Define the application

 

The first and most important step is to define the application. Smart cameras are designed for specific tasks, such as identification, presence checks, quality control, and measurement technology. Each of these tasks has different requirements in terms of sensor resolution, speed, and software functions.

Step 2: Specify the resolution and speed.

 

The resolution and speed are key performance criteria for smart cameras. These must match the size of the field of view, the smallest recognisable structure, and the processing speed.

 

  • Resolution: The larger the field of view and the smaller the detail, the more pixels are needed. For simple presence checks, a few pixels per feature are often sufficient, but significantly higher resolutions are required for reading 2D codes or text (OCR). Rule of thumb: The smallest relevant detail should be captured with at least three pixels, and more for text. You can use the product filter to narrow down the required resolution in megapixels or specify the required resolution directly via the horizontal and vertical pixel counts.
  • Frame rate: The camera must be able to capture and process images quickly enough to keep up with your production. High cycle rates or fast movements require short exposure times and a high frame rate to avoid motion blur. The maximum frame rate (Hz) can be selected using the product filter.

Guideline values: approx. 2–3 pixels per feature (presence), ≈ 4 pixels per module for DataMatrix/2D codes, ≈ 16 pixels character height for OCR.

Note movement: The exposure time must be short enough that the object travels < 1 pixel during the exposure; if necessary, use strobe lighting.

Shutter (global vs. rolling): For moving parts, prefer a global shutter (distortion-free). Only use a rolling shutter when stationary or with very short, strobe lighting. (Data sheet criterion)

 

 

Step 3: Consider the optics

 

Optics are a crucial part of the system. Even the most powerful camera will only deliver good results if the lens and lighting are correct.

 

  • Lens mount: Use the product filter to specify whether you need a C-mount, an M12 mount, or another type of mount. The C-mount is the industrial standard and offers the widest selection of high-quality lenses. The M12 mount is more compact but is better suited to simple tasks and limited installation space.
  • Lens selection: Selecting the correct focal length ensures that your test object is imaged on the sensor at the correct scale. The depth of field and optical resolution (MTF) of the lens in relation to the sensor pixel size are equally important; otherwise, a high megapixel count will result in blurry images. Practical tip: Derive the focal length from the sensor size and the desired field of view (FoV).
  • Lighting: Lighting significantly impacts contrast. Select lighting based on the surface and task: use ring lights for uniform illumination, transmitted light for contours and diffuse light for shiny surfaces. Without coordinated lighting, many tasks cannot be performed reliably. 
  • Trigger/timing: Check whether the smart camera offers strobe outputs for lighting control and IO/encoder inputs for precise timing. 

Step 4: Consider the installation environment and protection class

 

The camera must be suitable for the physical condition of your system. Therefore, check the following:

 

  • Size and weight: important for confined spaces or moving units, such as robot arms.
  • Protection class: Depending on the presence of dust, moisture, or cleaning processes, you will require a camera with the appropriate IP certification. IP65/67 protects against dust and water jets.
  • Temperature and vibration resistance: In harsh environments such as pressing plants or food packaging lines, robust housings and proven resistance are essential.

You can search directly for the protection class via the IP rating in the filter. The Dimensions and Weight fields also allow you to narrow down your search by size and weight.

 

Step 5: Check computing power and software platform

 

The tools available on smart cameras can differ (e.g. 1D/2D codes, OCR, matching and AI modules, if applicable), as can their performance at your target resolution/frame rate. In practice, this combination is more important than the CPU itself.

 

  • Platform/performance: Select a model that can reliably perform your inspection steps in real time at the required resolution and frame rate. If necessary, request performance specifications or test with sample parts.
  • Tool set & usability: Check whether the required tools are available and how easy they are to configure (graphical user interface versus programming). Closed platforms offer quick setup, while open platforms offer more flexibility.
  • CPU architecture (indicator): The architecture (ARM/x86) can provide an indication of performance and openness, but this is secondary to the toolset and actual performance.

Step 6: Clarify spectral requirements

 

Not all tasks can be completed using the visible spectrum. Therefore, check whether there are any special requirements.

 

  • Monochrome cameras: Ideal for capturing high sensitivity, good contrast and sharp details when colour is not important.
  • Colour cameras: Necessary when colour differences or colour codes need to be detected.
  • Special areas: Near-infrared (NIR), short-wave infrared (SWIR) or ultraviolet (UV) cameras are required to make certain materials visible or analyse special surfaces. While these cameras are more expensive, they open up additional inspection possibilities. Please note that SWIR/UV variants are model-dependent and rare in smart cameras. Consider alternative architecture if necessary.

Frequently asked questions about smart cameras

What is the difference between a smart camera and a traditional industrial camera?

 

A smart camera combines an image sensor, a processor and evaluation software in one compact device, delivering results directly to the control system. In contrast, a classic industrial camera only outputs image data and requires an external PC or vision system for processing. Smart cameras are therefore ideal for clearly defined tasks that require minimal integration, whereas industrial cameras offer greater flexibility and performance in complex scenarios.

Which industries are smart cameras particularly suited to?

 

They are used wherever compact, stand-alone solutions are required. Examples include the automotive industry (checking the presence of clips or screws), electronics manufacturing (reading codes and OCR), and packaging and logistics applications (checking completeness and position). In all cases, users benefit from easy integration and minimal space requirements.


How easy is it to integrate smart cameras into existing systems?

 

Many smart cameras are designed for high cycle rates and can perform inspections in milliseconds. The optimal combination of resolution, frame rate and processor performance is essential. This is usually sufficient for applications such as code reading on fast conveyor belts, but specialised high-speed solutions are necessary for extreme speeds or very detailed inspections.


What additional components are required?

 

In addition to the camera itself, most projects require a suitable lens and coordinated lighting. Lighting is particularly important as it significantly impacts contrast and image quality. Depending on the application, mounts, protective housings or encoder triggers may also be needed to ensure consistent results in the production environment.


Can smart cameras be operated without an external PC?

 

Yes, this is one of their biggest advantages. They capture and evaluate images directly within the device, only passing the inspection result on to the control system. However, it is important that the tasks are suited to the camera's computing power and tool sets. For very complex inspections or AI-based applications, a vision system with an external computing unit is often a better option.


How are smart cameras programmed or configured?

 

Many smart cameras can be configured via graphical interfaces, with drag-and-drop tools available for standard tasks such as barcode or OCR reading. More demanding applications require open platforms or SDKs that allow for greater customisation. This gives users the option of choosing between a quick setup and maximum flexibility.


Are there any plug-and-play solutions for common inspection tasks?

 

Yes, many smart cameras offer ready-made tools for standard tasks such as reading 1D/2D codes, optical character recognition (OCR) and simple presence checks. These functions can usually be set up without any programming knowledge and are ready to use immediately. However, to ensure reliable results, parameters such as lighting, tolerances and triggering must be tailored to the specific application.


How flexible are smart cameras when it comes to reprogramming for different tasks?

 

Most smart cameras allow you to store different 'jobs' or inspection programmes and switch between them as needed. This enables different products or variants to be inspected using the same camera. Depending on the platform, the range extends from simple job switching to fully programmable applications.


When is it preferable to use a complete vision system rather than a smart camera?

 

A vision system is the better choice when multiple cameras need to be synchronised, when very high resolutions (>5 MP) are required, or when complex algorithms are used. Smart cameras also have limitations when it comes to 3D inspections or AI/deep learning applications. In such cases, a modular system offers significantly greater performance and flexibility.


Can smart cameras be used for robot vision or pick-and-place applications?

 

Yes, they are well suited to simple robot vision tasks. They can detect positions and read codes, passing location information directly to the robot system. However, highly dynamic movements and complex 3D gripping processes require more powerful systems with greater computing capacity.


Are smart cameras suitable for high-speed quality control?

 

Many smart cameras are designed for high cycle rates and can perform inspections in milliseconds. The optimal combination of resolution, frame rate and processor performance is essential. This is usually sufficient for applications such as code reading on fast conveyor belts, but specialised high-speed solutions are necessary for extreme speeds or very detailed inspections.

Contact Tech News