How to design a robust GigE Vision Camera System

Designing a Robust GigE Vision Camera System

The adoption of Gigabit Ethernet (GigE) by the machine vision community as a camera interface and the development of the GigE Vision standard has enabled our industry to design, develop and deploy low cost, flexible and high-speed imaging solutions using standard PC architecture. GigE cameras offer a number of advantages:

• High Speed: GigE Vision cameras began by supporting 1GigE, but have adopted faster GigE topologies as the IT industry has developed them for general computing applications. These include 2.5GigE, 5GigE, 10GigE, 25GigE and beyond in the future. High Speed GigE enables high resolution, fast cameras to be used and built-in to robust high-performance vision systems.
• Cable Lengths: One of the key advantages of GigE cameras is there ability to transmit over long cable lengths. It enables the camera to be separated from the processor and for a camera network to be spread over large areas as required for example in some processing plants. This gives the ability for the data capture and image processing to be done in a centralised and possibly protected environment. GigE enables data transmission over distances up to 100 meters with standard copper Ethernet cables, and much further, spanning kilometres, when using network switches or fibre optics.
• Versatility and Scalability: GigE Vision is a networked architecture where multiple camera devices can use the same cables and the same network interface chipsets. The built-in contention mechanisms makes vision system integration simple and cost-effective for developers. As long as rules are followed, a system can be scaled and expanded from a single camera to a complex, multi-camera setup with ease. GigE is well-suited to both small and large-scale operations.
• Robust Data Transmission: GigE Vision has built-in recovery mechanisms that enable it to retrieve lost data ensuring reliable transmission even under harsh industrial conditions. Most applications require this level of reliability and redundancy.

Tips for ensuring system reliability and robustness
GigE Vision Camera systems will run consistently and without error when designed correctly and should be a priority in the design of a reliable system. Industrial, Scientific, Medical, Pharmaceutical, Defence and many other applications rely on the robustness of the vision system they use. Errors can have significant and far-reaching effects. GigE Packet Loss needs to be avoided and this requires good design and for the system to be built with real-time and ongoing data quality validation.

Here are three important aspects to designing and maintaining a reliable GigE Vision Camera System.

• Continuous Monitoring: Ensuring a close eye is regularly kept on system performance allows users to minimise or prevent issues like retransmission or packet loss.
• Leveraging Trigger To Image Reliability (T2IR): T2IR is an acronym used by Teledyne Dalsa to describe a framework they have developed to monitor and control the entire data transmission process from camera to computer host memory. It is an excellent example of how best to produce a high quality reliable system. T2IR manages in-built camera memory to store images temporarily in overflow conditions due to unexpected transfer bottlenecks due to any number of factors including network contention. More than that, it allows the system designer to monitor, audit and debug image flow during system integration. That means no lost data, no false data, and a clear source to identify and track any errors.
• Maintaining Data Integrity: Ensure that data transmitted across systems remains accurate and reliable. This protective approach is crucial for preserving the quality and consistency of imaging data, which is fundamental to system robustness.

Even a small packet loss, just 1–2% of retransmissions per frame, can be a sign of system instability, pointing to potential problems in system configuration or network performance. Left unaddressed, this can lead to significant declines in data quality and overall system efficiency. Frameworks like Teledyne’s T2IR should be purpose-built to identify, diagnose, and resolve these issues quickly, helping you restore system reliability and maintain optimal performance.

Overview of Topologies
GigE Vision Camera Systems can be configured in several topologies, each suited to different application requirements:

• Single Camera + Direct Host Connection: This is where a single camera connects directly to network interface chipset on adapter card or on computer motherboard. In instances where minimal latency and high reliability is most important this method works well.
• Single Camera Connected via Network Switch: In this setup, a single camera connects to the host computer via a network switch. This topology is ideal for systems that need parallel processing across multiple PCs or for system backup through multicast, providing both operational efficiency and redundancy.
• Multiple Cameras with Direct adapter Connection: Multiple network interface chipsets on adaptors or on computer motherboard are used to connect multiple cameras to a single host. This configuration is well-suited for synchronised data capture from various angles, enabling accurate and coordinated imaging across multiple viewpoints.
• Multiple Cameras Connected via Network Switch: Users can connect multiple cameras to a host computer with one or more network switches. This method is typically used in more complex set ups when scalability and flexibility are essential, allowing integration of numerous cameras within a robust network infrastructure. Additionally, this topology is best suited when IEEE 1588 (PTP) is required in the application.

Crucial Components for Each Topology
For each topology, certain components are critical to achieving and maintaining system robustness. The table below outlines the importance of these components, providing a quick reference for where to focus optimization efforts:

Legend:
★☆☆ (1 Star - Low Importance): Optional for basic operation; useful for improving performance but not essential.
★★☆ (2 Stars - Moderate Importance): Important for achieving good performance and reliability; recommended in most cases.
★★★ (3 Stars - High Importance): Critical for optimal operation; necessary for avoiding significant issues or failures.
N/A: Not applicable or not needed for this specific topology.

GigE Vision Camera System Topology Components

Optimisation in Focus
Each topology has its own unique optimisation requirements to ensure system robustness. Some configurations may call for advanced network setups and high-quality components, while others might benefit more from targeted software optimisation to address specific challenges.

GigE Vision Camera Systems Use Cases
To optimise GigE Vision Camera Systems, it’s important to understand how these various topologies and their core components match the demands of specific applications. These systems, whether built around a single camera or a multi-camera setup, can be adapted for everything from straightforward monitoring tasks to complex, high-speed, high-resolution imaging. The choice of configuration should be guided by the unique bandwidth demands and image acquisition rates of the application, as these parameters play a critical role in shaping both the system architecture and optimisation strategy.

Low Bandwidth Use Cases
Low bandwidth scenarios are ideal for applications with low to moderate data transfer rates, typically under 500 MB/s. These setups prioritise simplicity and cost-effectiveness, making them well-suited to tasks where high-speed data transmission is not critical.

• Single Camera: For applications such as basic quality inspection or straightforward surveillance, a 1GigE camera is typically sufficient. These systems usually operate with 2 to 5 MP resolution at frame rates between 30 and 60 FPS. The network setup is simple, requiring minimal software tuning.
• Multiple Cameras: In low bandwidth environments with multiple cameras, like basic multi-angle inspections or monitoring several conveyor belts, multiple 1GigE cameras can be employed. The emphasis remains on ease of setup and operation, with stable data flow and little need for complex integration or optimisation.

High Bandwidth Use Cases
High bandwidth applications demand substantial data throughput—often above 500 MB/s and sometimes into several gigabytes per second. These are essential for capturing high-resolution images and supporting detailed analysis.

• Single Camera: Use cases such as advanced inspection systems or medical imaging often utilise 5GigE or 10GigE cameras. These systems support resolutions above 5 MP with frame rates from 60 to 120 FPS. A robust network infrastructure and advanced software optimisation are necessary to manage the high data volume effectively.
• Multiple Cameras: For complex setups, such as multi-camera systems in scientific research or precision manufacturing, multiple 5 or 10GigE cameras are deployed. These configurations require an optimised, high-performance network and sophisticated software solutions to handle simultaneous high-resolution streams and ensure system reliability.

High Frequency Use Cases
High frequency use cases are defined by their need for very high line rates (1D line scan camera), high profile rates (3D camera), or frame rates often exceeding 1000 FPS (2D area scan camera). These setups are crucial for capturing rapid events, where real-time processing and data transfer are paramount.

• Single Camera: For high frequency applications like motion capture or rapid production line monitoring, 1GigE or 5GigE cameras are often used. These systems depend on a fast, responsive network and critical software optimisation to preserve data integrity at very high frame rates.
• Multiple Cameras: When multiple high-speed cameras are needed, such as in synchronised inspections or detailed motion analysis, 5GigE or 10GigE cameras are required. These systems demand complex networking and the highest level of software tuning to ensure synchronisation and manage large volumes of real-time data efficiently.

Summary of Use Cases
The table below summarizes the key features and considerations for both single and multiple camera system use cases. It directly compares various setups to help select the optimal configuration for an application.

When designing a GigE Vision Camera System to meet a wide range of industrial and scientific applications, each step is critical to achieving best performance. From understanding the importance of system robustness and selecting the right topology, to configuring your system to handle specific bandwidth and frequency requirements.

Key Takeaways:

• System Robustness:  With tools like Teledyne’s T2IR framework, you can monitor and maintain system reliability, preventing packet loss and ensuring data integrity.
• Topology Selection: The choice of topology—whether a single camera directly connected to an adapter, or multiple cameras integrated via network switches—has a significant impact on system performance. Understanding the strengths and limitations of each topology helps in configuring a system that meets the specific needs.
• Tailored Configurations: By aligning system configurations with the specific bandwidth and frequency demands of the application, the GigE Vision Camera System is both efficient and effective. Whether it is a low bandwidth setup for simple tasks or a high frequency configuration for capturing rapid events, the right setup is key to success.

Ask Adept Turnkey for Help
Networking GigE cameras can be complex. The selection and configuration of networking hardware and topology for GigE Vision is very specific to GigE Vision and requires considerable knowledge and experience. Our Vision Engineers at Adept Turnkey are well practised at designing, configuring and troubleshooting GigE Vision networks. Please click here and get in touch, if you need advice or help.