For image processing inspections, reliably capturing images of the object of inspection is more important than anything, and, like inspection systems, the setup of cameras, lenses, and lighting is essential.
Here, we introduce each representative type as well as how to make selections.

Cameras

Area sensors and line sensors

	Features	During image capturing
Area sensors	This is the most commonly used image-capturing method for image processing inspections during which two-dimensional images are captured. The image size is determined by the camera.	The objects of inspection are generally stopped, but using strobe lighting makes it possible to capture images of moving objects of inspection.
Line sensors	This image-capturing method is based on the same principle as a scanner or copy machine and involves continuously capturing a one-dimensional image (one line) to synthesize a two-dimensional image. This method is effective for continuously capturing images of objects of inspections bigger than those handled by area sensors.	The camera or object of inspection must be moving in one direction.

※You can scroll horizontally.

Monochrome cameras and color cameras

	Features
Monochrome cameras	With color cameras, because there are limits related to chromatic aberration and lighting colors, monochrome cameras are used for image processing in an overwhelming number of cases.
Color cameras	Color cameras are used to detect color discrepancies and judge color unevenness in objects of inspection. Because three times the amount of data is required than in the case of grayscale images of monochrome cameras, the image processing equipment must have high performance as well.

Resolution of a 1 CCD color camera

1CCD

The sort of array shown in the figure on the left is typical for a CCD color filter. The reason there is a lot of green is that human eyes are particularly sensitive to the color. Because the colors R, G, G, and B are assigned to each photodiode to record an image, the resolution falls to 1/4th the number of pixels compared to what it should be. Therefore, for each pixel, the nearby pixel output is used to perform interpolation operations and create RGB.

Lenses

Telecentric lenses and CCTV lens

Features of telecentric lenses

■ It is possible to capture images of objects of inspection that have depth from one direction

Example) If looking at the pins of a connector from directly above

Telecentric lens

An image of the tips
of all the pins can be captured

Non-telecentric lens

The sides of the pins are visible,
and there is variation.

■ Even if the distance from the object of inspection to the lens differs, there is no change in the scale factor (lens magnification).

Example) If capturing an image of an object that is at an oblique angle to the lens

Telecentric lens

An orthogonal image is captured
regardless of the WD difference

Non-telecentric lens

There is a sense of perspective due
to the WD difference.

Lighting

Selection of lighting suitable for an inspection

Because simply changing the lighting pattern can have a major effect on the contrast of an imported image, the key to successful image processing is said to lie in lighting technology. Careful consideration is therefore necessary during the initial lighting selection (the selection of a lighting method, wavelength, and shape suitable for the object of inspection).

Three selection points

• Lighting method
  The lighting method is selected based on the shape or unevenness of the object of inspection's front surface as well as what kind of inspection is to be performed.
  
• Lighting color (frequency)
  By changing the color of the object of inspection or brightening/darkening the color of the part you want to detect, it is possible to more strongly emphasize the part you want to see when capturing an image.
  
• Lighting shape
  The lighting shape is selected based on the size and shape of the object of inspection, the lighting method, and the installation location of the image processing inspection system.