Computer Vision

Image Processing pt.1


Image processing

Image processing describes an act of taking images such as a photographs or video frames and altering them using mathematical operations. A digital photograph is a set of arrays and matrices, and they each possess at least tree dimension. The smallest controllable element in a picture is called a pixel. In this day and age while talking about images we assume they are in digital form, but also analog and optical images can be processed. Firstly, being scanned and uploaded. By altering an image one can improve clarity, or remove noise, or extract the number, scale, or size of objects in a scene. Compressing images also falls in the category of image processing.


Fundamental steps to digital image processing
Description Example
First step Image acquisition Preprocessing Scanning, uploading, etc.
Second step Image Enhancement Highlight features of interest, and bring down obscured details. It is subjective Changing brightness, contrast, etc.
Third step Image Restoration Improving the image from an objective point of view. Using mathematical models to improve the quality.
Fourth step Color image processing Significant for using digital images on the web To change or optimice the color of an image.
Fifth step Wavelets and multiresolution processing To split the image in smaller parts for data compression and better use of the bandwith. Altering the wavelet coefficients, such as denoising, edge enhancement, etc.
Sixth step Compression Reducing the size of an image. Necessary to compress data in the uses of internet
Seventh step Morphological processing Dealing with software for showing image components Important for description and representation of shape
Eighth step Segmentation Converting an image into its smallest parts or objects Thresholding, color based segmentation, texture transformation, etc.
Ninth step Representation and description Raw pixel information to show the bounties of a region or all the points in the region itself Extracting information about quality, size and class of an object.
Tenth step Object recognition Assigning a label Labels that are based on the pictures description