Critical Dimension

The term Critical Dimension plays a significant role in the realm of imaging tasks, particularly when calculating the range at which an imager can perform the tasks of detection, recognition, and identification. The critical dimension generally refers to the smaller of the two principal dimensions, whether it is the width or the height, within an image. Understanding and accurately assessing the critical dimension is crucial for correctly interpreting the Johnson Criteria, a widely adopted standard for evaluating the performance of imaging systems based on the number of pixels covering a target in an image.

The Johnson Criteria establishes a framework for assessing the effectiveness of imaging tasks performed by an average observer using the given imagery. It does this by relating the number of pixels across a target to the probability that an observer can detect, recognize, or identify. The maximum range at which a certain number of pixels covers the target is said to be the range of the imager, to the given probability. By convention, most ranges are specified at the 50% probability level.

While actual targets are three-dimensional, the image of a target is two-dimensional. If the target were completely symmetrical, one target dimension could be used. However, practical targets such as humans are not uniform in width and height; therefore, only one of these dimensions, the width or height, can be used.

Because a human target is taller than it is wide, a longer-range value is estimated which poses a problem since the Johnson Criteria is calculated using the smaller of the image’s two principal dimensions for asymmetrical targets. An alternative method to determining the critical dimension is to take the square root of the area of the target if the target area is known. The “square root of the area” method is also useful for highly irregularly shaped targets.

When presented with performance range data, the customer should always ask what dimension was used as the critical dimension.