You are probably here because you would like to buy a machine vision system and want to understand the difference between CCD and CMOS image sensors. It is also more likely that you have heard the term CMOS more often. So, what is a CCD and a CMOS image sensor and how do they differ?
What are CCD imaging sensors?
CCD stands for Charged Coupled Device. It is a type of sensor used in digital and machine vision cameras to capture still and moving objects. The CCD sensor captures light and converts it into digital data to be converted into an image. CCD sensors were largely used by industrial machine vision systems for quality checks, inspection and control before the introduction of CMOS image sensors.
What are CMOS imaging sensors?
CMOS stands for Complementary Metal Oxide Semiconductor, which is a technology that powers integrated circuits. CMOS technology powers a lot of electrical devices of today including batteries, microprocessors, digital and smartphone cameras. CMOS sensors do not require special manufacturing techniques, unlike CCD sensors.
Difference between CCD and CMOS image sensors
Unlike the newer CMOS sensors, CCD sensors require special manufacturing which is often more expensive. As a result, CCD sensors are often very high-quality and light sensitive delivering crisp images with lesser noise.
CMOS sensors are cheaper to manufacture and use traditional manufacturing technology used for most microprocessors. They are also known to be better at power efficiency. According to the Handbook of Modern Sensors by Jacob Fraden, a CCD sensor can consume up to 100 times the power of a relative CMOS image sensor.
In a nutshell, you are likely to find a camera with a CCD imaging sensor where light sensitivity is an important factor or where high quality, high resolution images make all the difference.
Digging deeper: CCD vs. CMOS image sensors
Most production environments today use machine vision cameras powered by CMOS sensors because of their power efficiency and cost effectiveness. Technology has evolved over time and CMOS image sensors can now produce images that are less noisy. Each pixel on a CMOS sensor has its own C2V (charge-to-voltage) conversion, and each charge is therefore output digitally. Because of this, a CMOS sensor far exceeds at process speeds and power efficiency.
A CCD sensor has a limited number of nodes to convert each pixel charge on the sensor. This results in slower process speeds. However, since all the pixel can be devoted to light capture on a CCD structure, the output is crisp and brighter overall.
For machine vision, important KPIs are process speed and image noise. CMOS sensors allow for each pixel to convert digitally, resulting in a much lower bandwidth. Alternatively, high-speed CCD sensors are not as massively parallel as high-speed CMOS imagers. As a result, each CCD amplifier has higher bandwidth resulting in greater noise on the image produced. This is why most industrial applications use the latest CMOS sensors which can be designed to produce images with much less noise.
CMOS sensors and GigE Vision
Because machine vision and automation is highly dependent on speed, the EMVA introduces machine vision standards for imaging systems to comply with to ensure faster process speeds using the latest technologies. The GigE Vision is the latest standard for machine vision systems ensuring optimal imaging at very high speeds.
However, CMOS image sensors have gone even further to comply with 10 GigE machine vision technology. The 10 GigE Vision is the most advanced communications protocol, allowing for ten times the processing speed than its predecessor. Various CMOS sensors have already been used to manufacture 10 GigE machine vision systems, including the latest Sony Pregius CMOS image sensors.
Combined with the low noise, speedy and power-efficient performance of the CMOS imager, 10 GigE machine vision systems are performance powerhouses designed to cater imaging requirements in ultra-high-speed and fast-moving environments.
While CCD image sensors have been commercially available for a much longer period of time and are far more matured, rapid developments for CMOS sensors has made them the preferred choice for modern-day industrial machine vision.