Category: 

What Is Image Segmentation?

Article Details
  • Written By: Karize Uy
  • Edited By: Lauren Fritsky
  • Last Modified Date: 19 September 2016
  • Copyright Protected:
    2003-2016
    Conjecture Corporation
  • Print this Article
Free Widgets for your Site/Blog
Due to synthetic materials and furnishings, new homes burn about five times faster than those built 30 years ago.  more...

September 25 ,  1789 :  The US Bill of Rights was adopted.  more...

Image segmentation is a digital method that creates multiple layers and fragments of images from a simple image or picture. This technology greatly assists computers and machines in telling one object apart from another when scanning a one-dimensional image. In a picture, for instance, of a monkey clinging to a tree branch, image segmentation helps recognize and differentiate the monkey from the branch, making for an easier task in terms of image editing and recognition.

Generally, what image segmentation does is assign a value to each pixel, which are the small parts that make up an image. These pixels are then grouped together according to their likeness in areas like color, saturation, and proximity to each other. In this way, the image is then fragmented into different parts which technicians and digital editors can work with without having to alter the whole image, just the selected fragment. Many programs and software recognize the different fragments by highlighting the object when selected. Some programs even have the ability to isolate an object, then further isolate each of the object’s parts.

There are four commonly-used methods for image segmentation, the simplest of which is the threshold technique. Thresholding is usually for gray-scaled and black-and-white images, wherein the process assigns pixels only two possible values. Pixels recognized as background ones are assigned the value “0,” while object pixels are given the value of “1.” A colored image will turn into black and white when segmented by the thresholding technique.

Ad

Another method of image segmentation is the edge-based technique. This approach isolates images by distinguishing the outlines of each object, differentiating them from the background. This technique can be very effective for images with sharp contrasts, but is not as useful for blurry images and broken outlines. The region-based technique, on the other hand, not only isolates each object, but also isolates each regions of the particular object according to their characteristics. Many artists who use digital art often use this method for a more precise, but often meticulous, creation.

The most recent approach to image segmentation is the active contour model. This technique uses curved lines called “snakes” to make an object’s outline obvious. This is more effective for images with irregular shapes and outlines, as the snakes have the ability to conform automatically to the shape of the object. It is also used for noisy and grainy images that affect the vibrancy and color of the primary object.

Ad

You might also Like

Recommended

Discuss this Article

Post your comments

Post Anonymously

Login

username
password
forgot password?

Register

username
password
confirm
email