Industrial vision

Industrial visionRémi Gueudré2025-03-19T16:37:04+01:00

What is industrial vision?

Industrial vision applies computer-based control to production and research industries. It relies on a combination of hardware and software, providing operational guidance to detection tools. This guidance is based on image acquisition and processing to carry out their functions.

What are the uses of this solution? What about its components? What are the different types of this product?

The applications of this tool.

Industrial vision locates the object or checks the characteristic of the raw material within the sensor’s field using geometric recognition.

If the localization fails, the system cannot guide, identify, inspect, count, or measure the object, thereby compromising the success of the action.

PARTS LOCATION

The localization of material in a production environment is complex due to variations in appearance. These controls recognize parts based on their shape and dimensions, but these alterations, even in the most rigorous manufacturing process algorithms, can complicate precise and reliable marking.

THE GUIDANCE

Industrial vision determines the position and orientation of a part by comparing them to the required tolerances and validating the angle to ensure correct assembly.

Then, spatial guidance, in 2D or 3D, allows the robot or machine to locate and align the object.

IDENTIFICATION

This solution works for product labeling, particularly through OCR, which ensures the complex readability and compliance of barcodes and serial numbers: barcodes (1D), Data Matrix codes (2D), DPM codes, and printed characters.

OCR reads alphanumeric characters without prerequisites, while OCV verifies the correct presence of a string of characters and codes.

THE MEASURE

This detection measures the distance between points or geometric locations on an object and checks compliance with specifications.

In case of non-compliance, it sends a failure signal to the control system, activating a rejection mechanism to eject the object.

CIM ATLANTIQUE is at your service

For advice or any questions regarding the design, development, installation, and maintenance of your robotic installations…

Schedule an appointment now for more information on industrial vision, industrial robotics, and special-purpose machines:

Machine vision components

This detection system includes several key elements:

The lighting, which highlights the details to be analyzed.
The optics, which capture the image with precision.
The processing, which analyzes and interprets visual data to automate processes, ensure quality, and optimize performance.

THE LIGHTING

Lighting plays a crucial role in the performance of industrial vision solutions. It allows for image capture by analyzing the presence of light reflected by an object, rather than the object itself. The light source must be carefully positioned to optimize the quality of captured images and enhance the efficiency of the process.

THE OPTICS

Optics capture the image and transmit it to the camera sensor. The quality and price of optics vary, directly influencing the resolution of the obtained image. There are two types of optics: interchangeable and fixed. Interchangeable optics typically use C or CS mounts.

THE PROCESSING

Processing involves extracting source information from a digital image, including technologies like OCR for printed character recognition. It can be done via an external PC or standalone data software. This process includes several steps:

Image acquisition
Feature localization
mesures
Comparison with specifications
Decision-making based on the obtained results

The different types of this tool

There are three main categories of detection solutions.

1D vision processes

The 1D tool analyzes a digital signal line by line, rather than the entire image. It compares, for example, the differences between ten recently acquired lines and a previous group. This method is used to identify the presence of and classify defects on processed materials such as paper, metal, plastic, and non-woven rolls.

machine vision integrator

Matrix and linear systems

Linear systems offer advantages over matrix systems, particularly for inspecting cylindrical products. Instead of using multiple matrix cameras, rotating the part in front of a single linear sensor allows the entire surface to be captured. They also adapt better to confined spaces in industrial environments.

3D vision processes

3D inspections use multiple cameras or displacement laser sensors. In robotic guidance applications, this 3D tool informs the robot of the product’s orientation. These tools consist of sensors placed at different points, enabling precise triangulation on an object in a 3D space to locate objects.

Deep Capture, machine vision software

Deep Capture is software developed by CIM Atlantique, which uses advanced algorithms and deep learning to enhance the potential applications of industrial vision.

ARTIFICIAL INTELLIGENCE

Industrial vision based on deep learning, through our Deep Capture software, enables generalized detection via supervised learning. This approach improves the precision and efficiency of tools by recognizing various patterns and anomalies, thereby optimizing the performance of automated inspection tasks.

REAL-TIME ANALYSIS

Our software enables near real-time detection, achieving a performance of approximately 60 frames per second (fps) per camera. This speed ensures efficient and smooth analysis, ideal for applications requiring high precision and responsiveness in industrial environments

BEST PERFORMANCE

With our deep learning software, we ensure the best balance between the true positive detection rate (defects, objects to identify) and the false positive rate (rejection errors). This optimization guarantees precise and efficient inspection, reducing sources of errors and maximizing productivity.

SIMPLE AND INTUITIVE

With our deep learning software, the changeover times, whether mechanical or software, are virtually non-existent. This allows for increased flexibility, significant time savings, and rapid adaptation to production variations without prolonged interruptions to the industrial process.

TURN-KEY

Our specialized engineers handle all phases of software development, from design to commissioning. They ensure optimal integration of solutions tailored to your needs, guaranteeing performance and efficiency in automated production process algorithms.

Learn more about the uses of vision.

DEEP LIGHTING: THE REVOLUTION OF INDUSTRIAL VISION INSPECTION AND REFLECTION MANAGEMENT 2.0

We are excited to introduce DEEP LIGHTING, our advanced industrial vision inspection solution designed to address specularity issues on highly reflective surfaces such as lacquered, chrome, steel, and stainless steel parts. Discover how this innovation can transform your manufacturing processes

FAQ

What is the composition of an industrial vision approach?

An industrial vision system consists of lighting, optics, and image processing. It can include 3D cameras or sensors, deep learning algorithms for analysis, and efficient software ensuring rapid and accurate detection, often in real-time, with minimal adjustments.

What are the advantages of industrial vision?

Industrial vision offers several advantages: it enables rapid and accurate real-time inspection, reduces human errors, improves production quality, and optimizes process efficiency. Through supervised deep learning, it ensures precise defect detection and easily adapts to changes without prolonged manual interventions.

Who uses industrial vision?

Our industrial vision solutions are used by many companies in various sectors, such as automotive, electronics, food processing, and pharmaceuticals. These technologies enable the automation of inspection, quality control, and robotic guidance, thereby optimizing productivity, accuracy, and reliability of industrial processes while reducing costs.

What are the possible applications of vision solutions?

Industrial vision systems offer numerous applications: visual inspection to detect defects, automated quality control, reading barcodes and QR codes, robotic guidance for assembly or positioning, sorting and classification of products, dimensional measurement, compliance verification, and traceability of parts. They adapt to various industrial sectors, ensuring precision and efficiency.