JPEG-to-Vector Graphics

Understanding the process of converting JPEG images to vector images involves knowing the reasons for performing this conversion. It also involves understanding the broad types of methods for performing the conversion, and the steps for each type. Once you understand these aspects of JPEG to vector conversions, you'll be better able to perform such conversions yourself, and understand why a particular conversion resulted the way it did.

Reason

People convert JPEGs to vector graphics for different reasons. One reason is picture quality. Vector graphics can be enlarged to any size without loss of quality, whereas JPEGs degrade in quality when you increase their size. You can verify this by using your Web browser's magnification tool to zoom into any JPEG image. You may not see the individual pixels of the JPEG, but you'll be able to see the ragged edges of outlines of objects in the image. These ragged edges are telltale signs of the JPEG's pixels.

Conversion Categories

The process for converting JPEGs to vector graphics varies, but it falls into two broad categories, automated conversion and conversion by hand. Automated conversion involves feeding a JPEG to a stand-alone utility whose only purpose is to perform such conversions. Or, it involves running a command that performs such conversions from within a vector drawing program. Programs with such a command include Illustrator, CorelDraw and Inkscape. This conversion process is imperfect, due to a number of factors.

Colors

One reason automated conversion from JPEG to vectors doesn't result in an image exactly like the original is that many JPEGs contain more color information than vectors can handle. A vector image consists of a set of shapes, where a shape is usually an image region with a single color. Since a JPEG can contain millions of colors, it can potentially require millions of separate vector curves, a quantity that consumer computers can't handle. When a conversion process encounters such a JPEG, it therefore must reduce the number of colors before it performs the conversion. This color reduction will be noticeable for JPEGs with many colors. The resulting vector image will appear as if you ran a paint program's Posterize command on the JPEG.

Manual Conversion

The manual process for converting JPEGs to vectors usually takes much longer than an automated process. However, the result is a potentially cleaner image, and one that looks more like the original JPEG than what an automated process can provide. The reason for this is that human image recognition is better than machine-based image recognition, especially in the case of trained or experienced artists. A digital artist creating a vector graphic from a JPEG will ignore all the picayune details of the JPEG to focus on the image portions that best convey the essence of the image. For example, an artist tracing a vector from a JPEG image of a face can choose not to trace the moles, pores and other marks on the face. An automated process will mindlessly try to reproduce these features. To perform this conversion process manually, an artist will sample a color from the JPEG by using a tool whose icon is often shaped like an eyedropper, then she'll run her program's "Pen" tool, which produces vector shapes. She'll then trace over the outlines in the JPEG image to create these shapes.

References

• "Fundamentals of Computer Graphics"; by Peter Shirley; 2009