How to choose the right computer vision camera for my application?

One of the most important components of an artificial vision project is the election of an adequate artificial vision camera and its respective optics. The quality and the characteristics of the camera should be adequate for the application that is going to be done, and they have to guarantee an optimal performance of the capture. The capture speed of the image is also a feature to consider when developing the application.

Next, we tell you our suggestions for choosing an adequate artificial vision camera.

Camera characteristics

To guarantee that an adequate artificial vision camera is chosen, you have to define some features determined by vision system aspects as:

• Type of image: The type of image that we want to perform the application decreases the number of camera options to choose from. We have to consider a colour image or a grayscale photo and the resolution that it should have. The space and way of storing images are factors that could have an impact since they can influence the quantity of data that we pretend to store, something that directly affects the election of the quality of the images and their format.

• Capture speed: Capture speed or “frames per second” that the camera can capture, fps, is an important aspect in real-time applications, such as prosecution or object tracking, since they have to be processed in a short space of time.
• Connectivity: The camera must be compatible with other installation equipment, mainly, with the equipment in charge of image management. USB, Thunderbolt, and ethernet are some of the connectivity technologies that artificial vision cameras use.
• Type of sensor: The most used sensors are CMOS, sensor of active pixels, and CCD, a charge-coupled device. CMOS sensors are the most reclaimed in artificial vision applications because they have a lower electricity consumption, are cheaper, achieve better performance in low light conditions and can read a high number of pixels simultaneously.
• Sensibility to the light: The camera must have a high capacity to capture light to adapt to all kinds of atmospheres, even in those with low lighting.
• Integration in the application: It is necessary to check if our preference language is compatible with the camera and if there are libraries and drivers with which to control them. This feature has been simplified due to the creation of the standard Genicam, a generic programming interface that is being used in most industrial cameras such as Smart, GigE, IEEE and CameraLink.

Optics and compatibility of sectors

Optics is a main aspect when choosing a camera for artificial vision applications. The important points to take into account are:

• Compatibility with the camera: Optics must be able to capture the entire sensor field of the camera to avoid deformations, in addition to having enough resolution to the sensor itself.
• Lens quality: If we want to obtain a clear image without distortions, it is needed to use a high-quality lens.
• Opening: The opening of the objective influences the light quantity that enters the camera. A bigger opening allows the capture of images in low-light conditions, but it also affects the deep of the field.

Project characteristics

To be able to choose any equipment in any type of scope, the characteristics must be defined so that you should have as much as possible. These characteristics are determinants in choosing the equipment, optics and camera for these kinds of applications.

• Resolution: Quantity of details that can be observed in an image. The higher the resolution, the better the ability to detect small objects or details in the image. To measure applications, it is a very important point, to take into account.
• Focal distance: Distance from the sensor to the objects of which we want to take pictures. To obtain a quality image is important that the chosen optics be able to focus on the range where the object is located.
• Size and shape: The size and shape of the camera and the optics cannot be ignored when we have a limited space to do the application.
• Dynamic range: The field range of dynamic range is the capacity of the camera to capture details in the application areas that have different levels of lighting. A camera with a high dynamic range can capture very dark areas and very light without losing any detail.
• Field depth: Zone of the image that appears It is the range of distances in the scene that is clearly perceived in the last image. It is crucial to determine which areas of the scene will be focused and which ones will be unfocused
• Vision field: Area to capture in the image and in which are found the elements that you want to inspect through vision.

Points to take into account

The choice of a camera and optics adequate for artificial vision systems is crucial and it has a big importance since it can influence significantly in the development of the whole project, as we have been able to check.

• According to the application characteristics: The different application factors, such as the piece size to observe, are the main referent when choosing the equipment and they are determinants.
• According to the equipment compatibility: It is essential to consider that the camera, as well as the optics, possess complementary characteristics which contribute to meeting the application requirements to guarantee optimum and appropriate functioning. As well as a low low-quality resolution optics can be ineffective to a high-quality sensor, which can be inefficient too and even counterproductive for the acquisition of images.

Do you want to apply artificial vision in some of your projects? Contact with us!