What are the difficulties in the design of machine vision systems?

First: The Stability of Lighting

In industrial vision systems, applications are typically categorized into four main areas: positioning, measurement, inspection, and identification. Among these, measurement requires the highest level of lighting stability. Even a small change in illumination—such as 10-20%—can lead to a deviation of 1-2 pixels in the measurement results. This is not a software issue but rather a physical one, caused by changes in light that affect the edge detection in images. If not addressed at the system design stage, such issues can severely impact accuracy. To ensure reliable performance, ambient light interference must be minimized, and a stable, active light source should be used. Additionally, increasing the resolution of the camera hardware can also enhance accuracy and reduce sensitivity to environmental factors. For example, if the previous camera had a pixel size of 10 micrometers per pixel, upgrading to a higher-resolution camera with 5 micrometers per pixel effectively doubles the precision and improves resistance to external disturbances.

Second: Inconsistency in Workpiece Position

Whether it's an offline or online measurement system, the first challenge for fully automated testing equipment is locating the target object accurately. Even with mechanical fixtures, it's difficult to guarantee that the workpiece will always appear in the same position within the field of view. This is why precise positioning is essential. If the positioning is off, the measuring tool might not align correctly, leading to significant measurement errors. Ensuring consistent object placement is crucial for reliable and repeatable results.

Third: Calibration

High-precision measurement systems often require multiple calibration steps. These include optical distortion calibration (especially when not using a software lens), projection distortion calibration, which accounts for image distortion due to installation misalignment, and object space calibration, where the real-world size corresponding to each pixel is calculated. However, most current calibration algorithms are based on planar surfaces. When dealing with non-planar objects, special algorithms are required, as standard methods may not be sufficient. In some cases, custom calibration techniques are needed if no calibration plate is used, making it impossible to rely solely on existing software-based calibration solutions.

Fourth: Object Movement Speed

If the object being measured is moving, motion blur becomes a critical factor that affects image quality. The amount of blur is directly related to the object’s speed and the camera’s exposure time. This is a hardware limitation and cannot be corrected through software alone. Therefore, when designing vision systems for moving objects, it’s important to consider both the speed of the object and the capabilities of the imaging hardware.

Fifth: Software Measurement Accuracy

In measurement applications, the software can typically achieve an accuracy of about 1/2 to 1/4 of a pixel, ideally around 1/2. This is less precise than the 1/10 to 1/30 pixel accuracy seen in positioning tasks. The reason is that measurement software usually extracts only a few key feature points from the image, limiting its ability to detect fine details. As a result, achieving high precision in measurement requires careful system design, including proper lighting, accurate calibration, and suitable hardware.