QIA-64 is a suite of 85 Photoshop plug-ins for performing image processing and image analysis.

Notable features include:

• 8- and 16-bit image support in grayscale and RGB
• Brightness leveling and vignette correction
• Several Histogram Equalization techniques
• Easy access to other color spaces (Lab, HSV, HSL, YC_BC_R)
• Dual image arithmetic
• Color Fourier transforms and filtering
• Image enhancement tools well beyond Photoshop
• Color and grayscale morphology
• Binary image logic and processing
• Image measurements (area fraction, clustering, isotropic line lengths)
• Feature and Region data (more than 50 state-of-the-art parameters)
• Data logging and export for statistical analysis

Also included is a 58-page manual that integrates into the Photoshop > Help menu for easy access.

---

The tools in QIA-64 are organized in workflow-order under the Filter menu in Photoshop:
 

QIA-64 Workflow Steps
I. Adjustments.........................................	Adjustments are typically performed on images to correct acquisition problems (such as nonuniform illumination) or to convert an as-acquired image to a different color space in which the important subjects are more readily discriminated. These adjustments are generally performed before the enhancements described below.
II. Combinations....................................	Multiple images may be combined for various purposes: Gray scale or color images may be subtracted or ratioed for the purpose of leveling background or detecting object motion. Binary (thresholded) images may be combined with Boolean logic to define specific regions. One binary image may be used as a set of markers to select features in a second binary image In all cases, one image is first stored (in a special disk file) so that it can be combined with one or more other images. The stored image is not altered and remains in the disk file until replaced.
III. Fourier..........................................	Some processing operations are most straightforwardly applied in Fourier or frequency space rather than in the pixel domain. This space is particularly well suited to the removal of periodic noise (or conversely isolating just a periodic structure from random noise), and for finding matched or partially matched objects and shapes in a complex image.
IV. Enhancement..................................	There are many different algorithms that are applied to images in the spatial or pixel domain that deal with image noise, detail extraction (such as points, edges and lines), and conversion of one type of image to another that may be more suitable for thresholding and measurement (e.g., converting texture to brightness). The routines in this category are in addition to such built-in Photoshop Filter tools as the Gaussian Blur, Unsharp Mask, Median Filter, and Minimum and Maximum (which perform erosion and dilation, using a round or square neighborhood). QIA's additions are provided in two groups: Noise reduction or detail enhancement Extraction of specific components of the image, such as edges or texture.
V. Thresholding..................................	The thresholding tool in Photoshop simply divides pixels into black (foreground) or white (background) based on a user-adjusted value that is marked on the brightness histogram. The result is a binary (black and white) image in which pixels are either "on" or "off." The functions in QIA-64 have greater flexibility, including the use of color information. In most cases, image processing using the previous QIA-64 steps are required to prepare an image so that it can be successfully thresholded. The purpose is to assure that the pixels of interest (the foreground) have the same values wherever they happen to lie in the image, and that the values are unique and different from the pixels in the background. Thresholding is rarely perfect for every pixel, and the Binary Processing operations (shown in the next section) may be needed after thresholding.
VI. Binary Processing..........................	Operations performed on binary images are often necessary to clean up imperfect delineation of features and structures by thresholding. They are also important for isolating the points and lines that are most directly useful for subsequent measurement. These functions assume that the image consists only of black and white images. Any pixel that is not exactly white (value 255) and consequently part of the background, is treated as being black, or part of a feature. This allows features that have a range of brightness or color values to be measured.
VII. Image Measurements........................	These functions operate on the entire image area, or on a rectangular ROI that has been created. They are in the following categories: Image Content (Area Fraction, Image Brightness, Clustering, Total Line Length) Stereology Grids (five kinds) Histograms Ploting profiles of image data
VIII. Feature Measurements....................	A “feature” is defined in these functions as a group of contiguous non-white pixels that are completely surrounded by white background. Pixels are defined as touching and hence part of the same feature if they touch along sides or at corners (often called 8-connectivity). Depending on the nature of the image, features may represent objects, groups of objects, or other structures.