Very tough to map the small pieces of sentinel image with actual image or integrated image below the factors:

1.Complicated to identify the finite number of features over the integrated image with large size by supervised manner example entropy method

2. Identify the finite number of features over the each individual small pieces of images by supervised method

3. Need higher computational complexity to map between pieces of image and integrated image

4.Due to the image size factor the comparison accuracy would affect

Solution:

1.Better map the each individual piece of small image with other pieces of small images by using probability method

Procedure :

1. Find the finite number of features over the each individual small image by stand statistical operations

2. Use the probability method and map the each individual small piece of image with other small images.

3. There is a chance the comparison accuracy could increase

4. This procedure requires O( n2d) , where n is the number of small images and d is the number of features.