# What should come first-Radiometric calibration or co-registration?

For a pair of images,in the pre-processing of satellite images, which step should come first- Radiometric calibration or co-registration?

• radiometric calibration makes sense for one image but co-registration is for two or more images. you can first calibrate each image radiometrically and then co-register them to have the coordinates in the same reference system. each of these steps have different manners – Sepideh Abadpour Sep 11 '15 at 10:40
• @sepideh yes of course,co-registration is for a pair of image. I was asking that about a pair of images,which should come first? sorry to not clarify that in the question. – Nancy Sep 11 '15 at 11:07

Suppose that we have two images that we want to co-register or one image that we want to register to earth:

First step is to remove the errors in each image both geometrically and radiometrically. Each image has some geometric errors due to:

• Earth rotation
• Scan time skew
• Aspect ratio
• Panoramic effect (bowtie error)
• Earth curvature

These errors will cause the pixels to drift during image acquisition and so will effect radiometric information. So when we are removing (transferring the pixels to their correct position in image) these geometric errors, we should do radiometric interpolation, too. Radiometric interpolation can be done through:

• Bilinear interpolation
• Bicubic interpolation
• Nearest Neighbor

Also if the two images have different sizes, we should resize them in this step through the above interpolation techniques.

Second step is to co-register (determining the mathematical transformation between two image) the images. This can be done through different ways. One of them is to register both images to earth. When both images are registered to earth (the same reference system), they'll be coregistered to each other.
Different mathematical models are used for registeration based on different factors including the type of the sensor that is used to acquire the images. One of them that is used in HRSI images is terrain independent RPC coefficients

Thus we always remove the errors in each image first (calibration of each image) and then co-register them. This is true for all kinds of images in remote sensing including PolSAR, InSAR, Hyperspectral and Multispectral images.

Two images are co-registered when both of them are free of errors

The radiometric correction is necessary for the comparison of SAR images acquired with different sensors, or acquired from the same sensor but at different times, in different modes, or processed by different processors. Reference

Thus, radiometric correction/calibration needs to be performed before the co-registration of images.

with optical images, the radiometric calibration is used to convert the Digit Number into radiance. Basically, you can do it before or after the coregistration and it will not change much because it is a linear transform.

If you want to do change detection, you could then be interested in radiometric corrections. Those corrections can be absolute (if you have data on the atmospheric condition at the time of acquisition) or relative. In the relative case, you are normalizing the radiometric values of one image using the radiometric values of a second image. Therefore you need to spatially register first.

To conclude, most of the time it doesn't matter which is first, but some radiometric correction method rely on good registration, while registration method do not rely on calibration.

• But converting digital numbers to radiance is not a preprocessing step and also relative change detection. The question is about the preprocessing steps. I mean even in the above examples you should first do some radiometric calibrationbecause of the mensioned errors and then start to convert dn to radiance or in the second case you first calibrate each image, then coregister them and after that you start your main process through normalizing. Your examples is not about preprocessing calibration phase. They're about the main process either for change detection or radiance estimation – Sepideh Abadpour Sep 12 '15 at 4:11