Learn something new every day
More Info... by email
Automated image processing is a method by which images can be processed using pre-written, computer-based algorithms. The types of manipulations that can be achieved using automated processing of images include image segmentation, image filtering and image editing. As images become easier to collect with the technology in digital photography and digital image-based data collection, automated processing and image tool development pace the technological growth associated with image accumulation.
While many automated image processing algorithms are nothing more than a prerecorded macro in a computer program, techniques can be much more complicated, including making use of associated methods like machine learning and computer-based data processing. Automated image processing is often associated with machine learning, as computers are "taught" to search out certain image features and process the features according to the written program. As scientific data is often collected in the form of images, automated image processing is a necessary method by which scientists are able to quickly process large quantities of data.
Automated image processing software ranges in user interface ease and relative learning curves from data visualization and analysis programs to more straightforward image editing software. An intermediate user might make use of image processing to filter a set of images like digital photos — for example, for conversion of color digital images to a set of black and white pictures. More advanced users, or those who are interested in automated image processing for the sake of data analysis might use techniques that create an automatic workflow to segment images, count image artifacts or modify an image histogram.
Scientific data collection is largely based on the ability to make quantitative assessments from data sources that are often analog in nature, subjective, or more easily measured in qualitative measurements. Image processing algorithms allow scientists to quantify and compare images directly. Automated image processing increases the number of images a scientist can reasonably process, since a computer is able to process images rather than a scientist editing or taking data from images manually.
Limitations of automated image processing include an inability to account for image variations or outliers and the fact that computers are unable to process images and give a subjective critique of the final product. Many image editors are interested in creating quality images with filter effects or by removing unwanted information in the image. For most users, automated image processing means processing a set of images to make a single type of change again and again, allowing a computer to control the workflow. Computers are, however, incapable of making determinations about what is wanted and what isn’t, or about what “looks good.”
One of our editors will review your suggestion and make changes if warranted. Note that depending on the number of suggestions we receive, this can take anywhere from a few hours to a few days. Thank you for helping to improve wiseGEEK!