Debashish Saha
Debashish Saha

Reputation: 11

How to test on new images provided one training is done to the system

Hi this is the tutorial I am following.I have successfully applied to my problem statement also.and its successfully detecting what I want.But I want the same trained system to be used for detecting other images also.While detecting on other images I dont want system to get training again.Is it possible.I think so .I could see there is an XML file having data of training steps.Any help please. Thanks in adv

Upvotes: 0

Views: 16

Answers (1)

Liviu Stefan
Liviu Stefan

Reputation: 183

You need to address better your question. From what I understand you just trained a classifier to detect an object. Now you want to use the trained files to detect the object in other images.

det = vision.CascadeObjectDetector('nameOfYourTrainedFile.xml');
img = imread('image.jpg'); % load an instance of your object
bbox = step(det,img); % search for the object
obj = insertObjectAnnotation(img,'rectangle',bbox); % mark it
figure; imshow(obj); % display it

Upvotes: 1

Related Questions