The subjective interpretation of ISR imagery can lead to inaccurate methods of determining the accuracy to which differing ISR gathering equipment is capable of producing.
When engaging in Intelligence, Surveillance and Reconnaissance (ISR) activities, it is vital to be able to analyse information swiftly and accurately. However, the accuracy of which an image can be analysed is highly subjective.
Influences such as observer experience, image composition or even the physical ability of the person performing the analysis can all attribute to the accuracy in which an image is analysed.
For example, two analysts may interpret one image and propose two distinctly different opinions.
Johnson’s criteria divides the possible interpretation of an image into four categories;
Detection, Orientation, Recognition and Identification.
These categories are described below:
- Detection – An object is present
- Orientation – The object is symmetrical, asymmetric, horizontal or vertical
- Recognition – The type of object can be distinguished, a person versus a car
- Identification – A specific object can be distinguished, a woman versus a man, a specific car
The criteria express minimum required image resolution of a target in terms of line pairs or pixels for a 50 percent probability for an observer to differentiate between the categories and is shown below:
- Detection – (1 ± 0.25 line pairs OR 2 ± 0.5 pixels)
- Orientation – (1.4 ± 0.35 line pairs OR 2.8 ± 0.75 pixels)
- Recognition – (4 ± 0.8 line pairs OR 8 ± 1.6 pixels )
- Identification – (6.4± 1.5 line pairs OR 13 ± 3 pixels )
The following can be concluded from the image shown to the right:
- Detect – a group of objects can be Detected.
- Recognise – the objects can be Recognised as vehicles.
- Identify – the object can be Identified as a group of white vehicles and a red vehicle.
To use the Johnson Criteria to predict various sensor system performance, some key system performance specifications are required; such as sensor size and Horizontal Field of View. The size of the target is also required. Using the Johnson criteria for a selected category, an indicative distance can be calculated from the previously listed specifications.
Johnson’s Criteria is useful when attempting to objectively categorise the performance of visual ISR systems.