Sensor Image Quality - Terminologies that indicate it.

Updated: Dec 22, 2020

The scope of this article is to take you through certain terminologies that gives the indication of quality of the images that are captured with the sensor. Understanding these terminologies shall help with identification of the right sensor for the application.


The sensor datasheets primarily talk about -

1. Sensor Format

2. Pixel Size

3. Resolution

4. SNR

5. Responsivity

6. Dynamic Range


What do these terminologies indicate?


Sensor Format


We come across specification that talks about sensor format being 1/3” or 1/4" or 1/2.3” and so on. What does this indicate?


This number indicates the diagonal measurement of the active image area on the sensor. The inch referred to here is the optical form factor. 1” optical form factor is 16mm. This is derived from the times where we used tube cameras where image format of a tube placed in a 1” deflection coil was called 1” format.


The common formats that we come across are



Pixel Size


Pixel is the smallest element on the sensor that converts the energy from light (photons) to electrons which are then digitized on the sensor itself to provide us with an image. The size of each pixel indicates how many photons it can accept during the time the sensor is exposed to the light. The more the pixel size, the more light it accepts, thereby, providing us with better image.


However, we are seeing technology improving leaps and bounds to enable us to get higher quality images with lower pixel sizes. This improvements helps us to reduce the sensor size.

At this point, we would like to introduce you to sensor architectures such as FSI, BSI, BSI II.


They stand for:

1. FSI – Front Side Illumination

2. BSI – Back Side Illumination

3. BSI II – Back Side Illumination II

We shall cover this in depth in further sections.


Resolution


Resolution of the sensor indicates the number of the pixels that are present on the sensor. The number of pixels on the sensor is an indicator of the quality of the image to be expected.

However, quality of the image is subjective to the use case and application. For examples, for use cases where the camera would be used for performing OCR or wide angle image capture, then a higher resolution camera would be expected so that we can get more pixels to cover the scene. More the number of pixels, more the accuracy of the image.

On the other hand, the increase in resolution will impact the pixel size. For use cases where we need low light sensitivity or NIR sensitivity, we are looking at a higher pixel size sensor based cameras. This leads us to a lower resolution sensor. As technology improves on a daily basis, we are coming across high resolution sensors being able to give better sensitivity. We can discuss on this is the Responsivity section.


SNR


SNR stands for Signal to Noise Ratio. As the name indicates, it is the ratio of good signal caused by light falling on the sensor to the unwanted noise. We can consider this as one of the most important measurement of image quality. For each sensors, you can find the SNR curve that talks about the SNR compares of the photons per pixel. Considering the use case, we have to check the relevant part of the curve.


For example, for low light applications we have to check the part of the curve that indicates lower photos per pixel and read the corresponding SNR. When it comes to bright light application, we have to read the curve indicating higher photos per pixel and check the SNR corresponding to it. A camera good in low light need not be best in bright light.


Responsivity


Responsivity is a measure of the electrical output per optical input. The responsivity value indicates the sensors capability to product good images under low light conditions. The higher responsivity indicates that lesser photons are sufficient to create good electrical output.


The optical input is a function of wavelength. Thus, Responsivity curve or the Quantum Efficiency curve indicate the electrical output at for various wavelength of input light. Based on your use case, you shall have to check the curve at the necessary wavelength range.


Dynamic Range


Dynamic Range is a measure of how well a camera can reproduce details when both bright and dark areas are present.



In the above image, you can compare the image on the left side and the right side. We can notice the right image to have reproduced results with more details than the left image. The right image was captured with higher dynamic range camera when compared to the left image.

Based on your requirement, our solutions specialist at Regami, can work with your team in identifying the right imaging sensor for your application.


Please feel free to reach us if any queries


Thanks and Regards

Sarvesh Rajagopal

sarvesh@regami.solutions