A color-vision aid: take a photo, see where the red is. Optionally draw a bounding box to analyze a specific region. Histogram shows R/G/B intensity distribution.
A pixel is classified as red if all three conditions hold: R > G + threshold, R > B + threshold, and R > 60 (minimum brightness, so very dark pixels don't count). The threshold slider controls the first two — at threshold 35 a pixel needs R to exceed G and B by at least 35/255 to register as red.
This is deliberately simple. More sophisticated detectors use HSV color space (hue must be in the red range, saturation high, value not too low) which handles edge cases like pinks and dark browns better. The simple RGB version works well enough for most everyday red objects (apples, stop signs, brake lights, red fabric) and is easy to understand from the histogram.
The graph below the photo is a histogram, not a spectrogram. A spectrogram plots frequency over time (used for audio signals, like in the Magnetometer Waterfall demo). A histogram plots intensity distribution — for each value 0–255, how many pixels have that brightness in the red, green, and blue channels. Three overlapping lines, one per channel. This is the same view that's in every photo-editing program's color tools.