What is the process of converting a grayscale or color image into a binary…

Concept

Binarization (also called thresholding) is the image-processing operation that maps every pixel of a grayscale or colour image to one of exactly two intensity values, producing a binary image. A threshold value is chosen; each pixel above the threshold becomes one level (foreground/white) and each pixel at or below it becomes the other level (background/black). The defining property is the reduction of the intensity scale to just two possible values.

Application

The question describes converting a grayscale or colour image into an image that holds only two possible pixel intensity values. An operation whose output is restricted to two intensity levels is, by definition, binarization. Hence the process is Binarization.

Why the others do not fit

• Image Enhancement improves the visual quality or contrast of an image (sharpening, contrast stretching, noise removal) but keeps the full range of intensity values; it does not reduce the image to two levels.

• Grayscale Conversion turns a colour image into shades of gray, retaining many intensity levels (typically 256), not two.

• Quantization reduces the number of distinct intensity levels to a smaller finite set, but that set is generally more than two; reducing specifically to two levels is the special case named binarization.

• Segmentation partitions an image into meaningful regions or objects; the output is labelled regions, not a two-valued intensity image.

Cross-check

Standard image-processing libraries name this exact operation accordingly: MATLAB's imbinarize, scikit-image and OpenCV's thresholding all take a grayscale source and emit a two-valued (0/1 or 0/max) binary image, confirming the term Binarization.