You need to convert the image to a logical mask that describes the black regions. In order to do that, you need to define what "black" means in whatever context you're dealing with. There are existing examples of thresholding and color-based segmentation throughout the forum; otherwise, you'll have to provide an example image and a description of the expectations.
Once you have that logical mask, you can use bwarea(), nnz() or perhaps regionprops() to get the area in pixels. It depends if you want to find the total area or the area of each isolated region.
