Image Histogram

For each channel it counts how many pixels have each brightness level from 0 to 255 and groups the levels into the number of bins you choose: 16, 32, 64, 128 or 256, with 32 as the default. With 256 bins you get one count per level. The counts of each channel add up to the number of pixels.

Each channel gets its minimum, maximum, mean, median and standard deviation on the 0 to 255 scale. The image also gets an entropy estimate, a sharpness estimate and a flag that says whether it is fully opaque.

Look at the last and first bins. A tall last bin means many pixels are pure white in that channel, so detail in bright areas is lost, and a tall first bin means crushed shadows. A maximum below 255 or a minimum above 0 shows unused range that levels or the Image Normalize tool can stretch.

Entropy measures how varied the gray levels are: a flat color has 0 and a detailed photo is often above 7. Sharpness is the standard deviation of an edge filter on the gray image, so blurry or flat images score low and crisp, detailed ones score high.

Color images are measured as red, green and blue, grayscale images as a single gray channel, and an alpha channel is added when the image has transparency. CMYK and 16-bit images are converted to 8-bit sRGB first, so every value uses the 0 to 255 scale.

JPEG, PNG, GIF, WebP, TIFF and AVIF images up to 5 MB. Animated files are measured on their first frame.

Images larger than 4000 pixels on a side are scaled down to fit 4000 pixels first, and the reported width, height and pixel count describe that measured copy. Smaller images are measured at their full size.

The mean is the average level, while the median is the level that half of the pixels fall below. A small, very bright area raises the mean but barely moves the median, so comparing both shows whether a few extremes skew the image.

A low standard deviation means most pixels sit close to the mean, which points to a flat or low contrast image. A high value means the channel spans a wide range from dark to bright.

Yes. The result contains the histogram counts and statistics as structured numbers, which the MCP version of this tool returns directly to agents and scripts for checks such as rejecting underexposed uploads.