Blog_banner_beyond_the_datasheet

Comparison of CMOS (IMX174) Cameras

Posted by Gretchen Alper on Wed, May 20, 2015

With how good the latest industrial CMOS image sensors have gotten, it might seem that a camera just simply connects the image sensor to an interface. 

Thankfully for us this isn’t the case! The below images show two different IMX174 sensor-based cameras looking at the same scene (images from the Adimec camera are on the left, the second row of images shows the difference when zooming in to the scene).

For the first set of images, the color differences are obvious, although in a critical decision-making situation having a true representation of the colors as they are might not lead to too many mistakes.

In the second set of images, the image quality comparison is further apparent.  The lack of sharpness in the image from the competitor’s camera makes it difficult to read any text or distinguish the lines on the chart. There are also many false colors apparent. All these image artifacts reduce the usability for measurements and decision-making.

      Adimec IMX174 (CMOS) Camera

                           

Competitor IMX174 Camera

Adimec-imx174-cmos-camera-comparison

Adimec-imx174-cmos-camera-comparison-2

So what is different between cameras utilizing the same image sensor?

Despite the high quality of the starting image sensor, there is still significant image processing required.  The way that any processing is done can have a dramatic impact on the resulting image.  For example with color processing or de-bayering, a proper color matrix and interpolation and good white balance are necessary. This usually goes beyond the textbook implementations and depends on years of experience and unique real-time implementation algorithms. Adimec also does color processing over temperature to ensure the color quality is maintained regardless of conditions.

There are also other image sensor related image processing that required and should be calibrated over temperature, such as defect pixel correction, uniformity correction (column, row, shading), and linearity correction.  This is all done to optimize image quality and take full advantage of the MTF of the image sensor.

A good implementation with years of experience is critical to real-time decision-making.

For more information on the differences between cameras with the same image sensor, see our free ePaper

Topics: CCD vs. CMOS

Previous blog:

Next blog: