GENERATING REPRESENTATIVE SETS AND SUMMARIES FOR LARGE COLLECTION OF IMAGES: PRE-CLUSTER (1)

In pre-cluster we have used three different techniques which are as following.

RGB and Gray scale Techniques
Using these two pre-cluster techniques, we simply stored images in RGB form and applied clustering algorithm (k-means). For gray scale technique, we converted all the images in gray scale and stored in vectors and applied again the cluster algorithm in second phase. From the result of clusters we found that the pictures with same appearance with different pixel values don’t go to the same cluster. It means taking a picture of any particular landscape with normal mode and taking the same picture with zoom, would not go in the same clusters although both have same visual features. Because of this big drawback we can’t stand on finding a representative set and summarization which holds the selection criteria. So, we stopped our experiments with this pre-cluster stage and we focus only on the next two techniques.

Random Windows

As we mentioned earlier that the idea of cropping images initiated from n-gram model, however random windows are not suitable for the n-gram model. We initiated with taking n-windows of same size with certain fix coverage of each images from random pixel point of an image. More formally For each image of the dataset P and for the fix coverage C (i.e. window should cover some fixed portion of image like 85%, 75% and 66%), generate N random windows from (r_x, r_y) pixel point (i.e. 3 windows of 75% coverage of each images) where r_x and r_y is random point from where the image with fix coverage will be cropped. One can see from the figure 4 that the black dots are windows initial points which are totally random. After having the N windows set of original data set, we apply sift algorithm to generate and store descriptors of each random windows for the next stages. Descriptors are K-by-128 matrix, where each row gives an invariant descriptor for one of the K key points. The procedure is mentioned in algorithm 1 and the output can be seen in figure 5 with respect to the original image.

Fig4Generating Representative_decrypted
Figure 4 : Random and sequential windows

Algorithm 1: Random windows generation

1: input: original image set

2: output: N random windows of each images with certain coverage C.

3: set the coverage C and the number of windows N to generate random windows

4: for each image img do get the size of img [rows columns] = size(img) manipulate window width and height according to coverage C: w_width = C * column; w_height= C * rows;

5: for each image img, to generate N random windows do generate random pixel point (r_x,r_y):

crop the image img : cropped_img = imcrop(img, [r_x, r_y, w_width, w_height]) save cropped_img (window) at the output directory 6: end for 7: end for