Johnson's Criteria

Night vision systems enabled the measurement of visual thresholds following World War II. The 1950s also marked a notable development in the performance modeling of night vision imaging systems. Throughout 1957-1958, John Johnson, a [http://www.nvl.army.mil United States Army Night Vision Lab] (now called Night Vision & Electronic Sensors Directorate) scientist, was working to develop methods of predicting target detection, orientation, recognition, and identification. Working with volunteer observers, Mr. Johnson used image intensifier equipment to measure the volunteer observer's ability to identify scale model targets under various conditions. His experiments produced the first empirical data on perceptual thresholds that was expressed in terms of line pairs. A line pair is the distance subtended across parallel light and dark line's at the limit of the observer's visual acuity. A line pair is the equivalent of two pixels.

At the first Night Vision Image Intensifier Symposium in October 1958, Mr. Johnson presented his findings in a paper entitled Analysis of Image Forming Systems. The paper described both image and frequency domain approaches to analyzing the ability of observers to perform visual tasks using image intensifier technology. Later referred to as Johnson's Criteria, Mr. Johnson's findings proved to be such an important breakthrough in understanding the performance of visual devices that it would provide the basis guiding the development of future systems. Using Johnson's Criteria many predictive models for sensor technology have been developed that predict the performance of sensor systems under different environmental and operational conditions.

The minimum resolution according to Johnson's Criteria are:

* Detection - an object is present: 2 +1/-0.5 pixels
* Orientation - symmetrical, asymmetric, horizontal or vertical: 2.8 +0.8/-0.4 pixels
* Recognition - the type object can be discerned, a person vs. a car: 8 +1.6/-0.4 pixels
* Identification - a specific object can be discerned, a woman vs. a man, the specific car: 12.8 +3.2/-2.8 pixels

These measurements give a 50% probability of an observer discriminating an object to the specified level.

Recent work at the NVESD has given rise to a replacement of the Johnson criteria. The Targeting Task Performance (TTP) metric has been shown to be far more robust in representing sensor performance.

References

*John Johnson, “Analysis of image forming systems,” in Image Intensifier Symposium, AD 220160 (Warfare Electrical Engineering Department, U.S. Army Research and Development Laboratories, Ft. Belvoir, Va., 1958), pp. 244–273.

*Harney, Robert. "Combat Systems vol. 1: Sensors" , Chapter 14: Imaging and Image-based Perception. 2005. http://www.nps.navy.mil/se/harney/cbt1ch14.pdf