matlab - Calculate precision and recall on WANG database -

i have made cbir system in matlab , have used similarity measurement euclidean distance. using each query image retrieve top 20 images.

i have used wang dataset testing system.
contains 10 classes(like african people, buses, roses etc.) each containing 100 images.(1000 images in total).

my method:
1. using correlogram, co-occurence matrix(ccm) , difference between pixel scan pattern(dbpsp) constructing vector(64+196+28=288 dimensions respectively).

2. each of 1000 db image have vector constructed beforehand.
3. now query image comes , construct it's vector too(228 dimensions again).
4. i use euclidean distance similarity , sort db image vectors in descending order of euclid distance.

5. top 20 results shown.

6. in 20 can have tp or fp.

for single query image can calculate precision , recall , plot pr-curve using link.

how can same whole class?

my approach: each image belonging class find top 20 images , it's respective tp(true positives) , fp (false positive).

        tp   fp  image1  17   3   image2  15   5   ...   ...   image100  10  10   total   1500 500   

precision of class =1500/(2000) = .75 (is right??)
recall of class ---> stuck ??
pr-curve ----> stuck ?? links said need classifier , not... confused.

so noted, can calculate precision follows.

p = tp ./ ( tp + fp ); 

however, need have either have fn or number of total falses calculate recall. discussed in chat, need find way determine fn , fp data. can use following formula calculate recall.

r = tp ./ ( tp + fn ) 

if have confusion matrix or data, can use custom confusionmat2f1.m calculate precision, recall, , f1 score. assumes confusion matrix formatted how matlab defines it. explanation of each line inline. please let me know if want more clarification.

function [f,p,r] = confusionmat2f1( c )     %% confusionmat2f1( c )     %     % inputs     % c - confusion matrix     %     % outputs     % f - f1 score column vector     % p - precision column vector     % r - recall column vector      %%       % confusion matrix probability     m = sum( c, 3 );      % calculate precision     p = diag(m) ./ sum(m,1)';      % calculate recall     r = diag(m) ./ sum(m,2);      % calculate f1 score     f = f1( p, r );