image - Combining 2 histogram data for a new Data in OpenCV -


I want to compare 2 images and if they are equal to placing 2 images. I will use the HSV histogram I calculate and compare the distance between the histogram.

Now when the third image is received, I have to compare it to image 1 and image 2 (already stored in an identical type of image).

The problem of comparing the above comparison is that the increase in images also increases computational cost

So what do I want to do it, if 2 images are equal I want to cluster, so the same features will be compared to the clustering features of similar images in the future.

> Option 1 If I merge 2 histograms then this would be correct? I do not think so, but I'm not sure?

Option 2 How to use 2 image feature distribution, I calculate 2 new additions from histogram as a combined distribution of both. ? Is this sound right?

I should take this step by step:

Functions: Comparison Do 2 pictures if they are the same then keep both, then merge in some way Feature Space: HSV Histogram.

Option 1 Is histogram merge right?

Yes, because you use the histogram and do not sign, you can divide only the two bins by histogram and two.

Tour: If you want to merge additional images, you will need to keep track of the number of merge already, so that you know how to weight

  example : Histogram with a bin, p1 = 2, p2 = 6, p3 = 10 merge p1, p2 to m2 2: (2 + 6) / 2 = 4 merge m_12 and p3: (weight * Value M2 2 + + (Vig * value P3)) / 2 = ((2/3 * 4) + (1/3 * 10)) / 2 = 6 [= (P1 + P2 + P3 ) / 3]   

tl; dr Yes you can merge them

Option 2 How about using 2 image feature distribution, I calculate 2 new additions from histogram as a combined distribution of both images . ? Is this sound right?

Yes, although I do not know immediately how you want to do it.

If you want to speed up your program, then you should check different distance measures (I remember SQFD and Earth Movers unfortunately for ATM signature distance). Often they have a fast but thick lower band. It can be used to make a good low bound for distance, so that you can reduce your search space.

Increase in images also increase computational cost.

Check hierarchical clustering to find favorable data structures for large numbers of images.

Comments

Popular posts from this blog

Java - Error: no suitable method found for add(int, java.lang.String) -

java - JPA TypedQuery: Parameter value element did not match expected type -

c++ - static template member variable has internal linkage but is not defined -