Built-in function for bags of visual encodings of words in python

Question

In matlabI have this feature

 function psi = encodeImage(encoder, im)

This function accepts

• imwhich is a list of image names

• encoderThat bovw.matI have this file as an encoder

This function performs a bag of visual words that encode and return spatial histograms of images.

I use these histograms for training in the SVM classifier.

I am performing this task in python and I do not want to implement a visual word coding package, since my main task is to implement SVM. Is there a built-in function in python that interferes with the coding of visual words and returns spatial histograms, so I can train the SVM classifier on histograms.

Answer 1

Are you doing something like this?

http://www.robots.ox.ac.uk/~vgg/practicals/category-recognition/index.html

There is a computer vision library called Vlfeat. This version of Matlab is under active development. However, there is a Python interface . Supports all the basic functions of image processing:

• Scale Invariant Function Conversion (SIFT)

• Tight SIFT (DSIFT)

• Integer k-value (IKM)

• Hierarchical Integer K-Tool (HIKM)

• Extremely Stable Extreme Regions (MSER)

• Quick image segment change

, Pyvlfeat . , . . , , . - , , .