I am doing my research in automatic image annotation. I tried by using vision toolbox, but able to solve only for one object. how to proceed for multiple objects

1 visualizzazione (ultimi 30 giorni)
Datti Nagadhara Harini
Datti Nagadhara Harini il 10 Apr 2014
Modificato: Image Analyst il 26 Lug 2014
my code for identifying cats in an image.
load('cats.mat');
imDir = fullfile(matlabroot,'toolbox','vision','visiondemos','cats');
addpath(imDir);
negativeFolder = fullfile(matlabroot,'toolbox','vision','visiondemos','non cats');
trainCascadeObjectDetector('stopSignDetector.xml', data, negativeFolder, 'FalseAlarmRate', 0.2, 'NumCascadeStages', 5);
trainCascadeObjectDetector('cats.xml', data, negativeFolder, 'FalseAlarmRate', 0.2, 'NumCascadeStages', 5);
img = imread('test1.jpg');
detector = vision.CascadeObjectDetector('cats.xml');
bbox = step(detector, img);
detectedImg = insertObjectAnnotation(img,'rectangle', bbox, 'cat');
figure; imshow(detectedImg);
rmpath(imDir);
  2 Commenti
Nomi khan
Nomi khan il 25 Lug 2014
Dear harini , im also working on AIA research, can u please share details of your work done with me at nouman.ali@live.com hope that we can help each other
Nomi khan
Nomi khan il 25 Lug 2014
Modificato: Nomi khan il 25 Lug 2014
u need to create a codebook for this , the technical names are codebook, bag of words, bag of images or visual dictionary

Accedi per commentare.

Accedi per rispondere a questa domanda.

Risposte (1)

Dima Lisin
Dima Lisin il 1 Mag 2014
You would need a separate detector for each object category.

Categorie

Scopri di più su Computer Vision Toolbox in Help Center e File Exchange

Tag

Prodotti

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by