Browse Prior Art Database

System and Method for Automatic Pattern Based Image Retouching

IP.com Disclosure Number: IPCOM000246026D
Publication Date: 2016-Apr-27
Document File: 8 page(s) / 628K

Publishing Venue

The IP.com Prior Art Database

Abstract

Currently, people has more and more requirement on image retouching and photoshop to get a beautiful photo. So there are many tools generated for this purpose. But a problem found in such tools is, people is hard to control the PS result, sometimes it will be far different than the original. Take human face retouching as an example, by different user, or even the same user to use current tools, the skin color, sizing of eyes, etc may be set differently everytime, the result is people in the photo looks like non-consistent. This disclosure proposes an automatic method for extracting features from user preferred image, saving as pattern, and then applying to other images in the future. The goal is to simplify the retouching process and get a consistent result.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 80% of the total text.

Page 01 of 8

System and Method for Automatic Pattern Based Image Retouching

The main idea of this disclosure is: user selects a preferred image as template, the system extracts features from this image, saves as patterns, and then apply to other images in the future.

The advantage of this method are: 1) Simplify the image photoshop process. 2) Keep consistent result after photoshop.


1. User select a preferred image as template;

1



Page 02 of 8


2. User select object which will be used to extract features;

2



Page 03 of 8


3. System extract features from the object, like color distribution, contrast distribution, size of facial features, etc. All of these feature extracting technologies are all typical signal processing and image processing technologies, they are not claim points of this disclosure;


4. System save the extracted features to a pattern and record in pattern database associated with the template image;

3



Page 04 of 8


5. User get a new image and pick out the pattern;

4



Page 05 of 8

5



Page 06 of 8


6. System apply the patterns to the new image. Here some pattern matching technologies need to be used such as Histogram Matching algorithm, they are not claim points of this disclosure. In the sample below, it shows applying the skin color pattern to the new image.

6



Page 07 of 8

7



Page 08 of 8

Still take human face as example:

One existing method is to give user several feature options, and then ask user to manually select ratio like size, color, etc for each of the feature...