The digital image sensor in the camera see a silhouette when a foreground-object is placed in front of a strongly lighted background. The foreground-object looks black. The contrast between the very bright background and the relatively dark foreground object is so large the image sensor, cannot resolve the detail of the darker object.
Our eyes are better than an image sensor at seeing the detail in contrasting light situations. However, very large differences mean we too see only a silhouette. For example we might see only the silhouette of a small toy held up against a very bright sunny sky. We will probably not be able to see the detail of the toy against the brightness of the sunny sky leaving it black to our eyes.
In photographic terms, more than two stops difference between foreground object and bright background will tend to create a silhouette. A brighter background or darker foreground object will deepen the blacks in the silhouette.
The contrast of two stops of light or more between silhouette object and bright background is a good guide to creating a silhouette. It is not exclusive. Some colours may still be resolved by the camera in the foreground (darker) object at two stops difference. The effectiveness of the camera light meter in determining tones may also be a factor. If you are metering the contrast between the two with a hand-held meter that may give a different result to the camera meter and so some testing might be necessary to verify a silhouette has been created. In reality a greater difference will be more likely to create deeper blacks in the silhouette.
As image sensors are improving all the time it may not be as easy to produce a silhouette in the future. The dynamic range of contrasts the camera can ‘see’ is increasing. As better sensors are released we will need a bigger difference than two stops light intensity between foreground object and background.