A linear camera captures a single line of pixels for each acquisition, unlike matrix camera which capture an entire two-dimensional image in a single instant (a complete grid of pixels that immediately represents the scene).

In practice, the linear sensor does not directly provide a complete photograph, but to obtain it, a sequence of one-dimensional lines must be captured. The image is then digitally reconstructed by concatenating many consecutive lines in the correct order. This process typically occurs on the processing system, which synchronizes the acquisition of the lines with the object’s movement (for example, via an encoder).

The result is that, unlike array cameras that return complete 2D images, line-scan cameras require an additional reconstruction step. However, this approach allows for extremely high resolution along the scanning direction and the inspection of very large or continuously moving surfaces, making linear scanning a fundamental technology industrial applications.

There are both models that produce black and white images and models that produce color images.

To acquire images with a linear camera, the movement of the object in the frame must be synchronized with the acquisition. This is done using an encoder.

The encoder converts mechanical movement into an electrical or digital signal, generating a periodic signal. This signal is interpreted by a controller to decide at what times to capture each individual line with the linear camera.

Speed and acquisition frequency must be set correctly to avoid artifacts in the image.

LThe scanning speed is limited by the camera’s maximum acquisition rate.

The maximum acquisition rate is limited by:

The camera can only acquire a new line after it has finished reading the previous line, so it is important to send acquisition triggers to the camera at a frequency lower than the camera’s maximum frequency.

If the trigger is lost, some lines are lost and therefore parts of the object in the frame may not be visible in the image or there may be artifacts.

There are lenses for industrial cameras designed to suit the specific needs of linear cameras.

They are designed to provide a wide field of view along the scanning direction.

Because they capture an image line by line, the amount of light hitting the sensor is relatively low in each individual acquisition. Line camera lenses are designed to concentrate light onto a single line, allowing for uniform illumination of the scene.

They enable extremely precise and continuous focusing on a single line, without distortion. This is particularly important for industrial applications where image quality and precision are crucial.

Linear cameras are particularly suitable for vision systems for quality control  that require scanning of moving objects (e.g., on a conveyor belt) or large objects.
Using linear cameras, it is possible to capture a single image of the entire object by capturing a sufficient number of lines: since they only have horizontal resolution, the vertical resolution of the image obtained is given by the number of lines captured.

In one of our automated inspection systems for PVC reels, a series of linear cameras are used to scan a strip of film 2000 mm wide and up to 3600 mm long, as the reel unwinds, and detect any defects.

Line cameras are also suitable for scanning circular or cylindrical objects. By moving the object in a circle on its axis and using line-by-line scanning, it is possible to obtain an “unrolled” image of the object’s circular surface, which can be processed for measurements or quality control.
For example, it is possible to scan the outside of a jar, obtaining a “flattened” image of the surface.

Linear cameras are also very useful in cases where the layout of the conveyor line for the objects to be scanned does not allow for positioning cameras or illuminators to frame or illuminate the entire scene.