Detection, recognition and identification (DRI) is a technique used in photography and imaging to enhance the visibility of details in light and dark areas of an image. DRI stands for detection, recognition and identification, and in thermal imaging drones, it is a key metric used to compare the performance of thermal imaging cameras.
What is DRI?
DRI is a set of generally accepted standards for measuring the distance at which a thermal sensor can produce an image of a specific target. These standards were originally developed by the US Army in the 1950s and are often used to compare the performance of imaging cameras. The standards have not been updated much to this day, so they continue to be used.
Detection
Detection is the distance at which the target first appears in the image. This "target" is something out of the ordinary, hotter or cooler than the surrounding environment. This will allow us to quickly detect abnormal information in the image, but not to determine what it is.
Specifically, it will be visible on at least two pixels, so there is not enough information to confirm what the target is at this distance, just that it is there.
Recognition
Contrary to expectations, recognition does not mean that you can recognize a person. Recognition is the distance at which you can determine the class of an object (is it a person, animal or vehicle). Typically, drone thermal imagers can quickly distinguish between three objects: cars, boats and people, and display corresponding icons.
Identification
Recognition is when you can distinguish different objects of the same class within a certain distance. For example, identifying the type of vehicle (truck, SUV or car) or identifying whether a person is a soldier or a civilian.
How is DRI calculated?
DRI is calculated using the Johnson criterion, which is based on the number of pixels on the target. According to the Johnson criterion, the minimum resolution required for an observer to achieve a 50% probability of identifying an object is:
Detection: Presence of an object – 2 pixels on the target
Recognition: Identify the type of object, is it a person or a car – 8 pixels on the target
Recognition: Identify a specific object, man vs. woman, car model – 12.8 pixels on the target
These calculations are based only on geometric parameters – target size, distance, lens focal length and camera detector pixel size. Signal level, detector sensitivity, atmospheric conditions and other factors are not considered. Weather is almost never ideal, and in actual use, these distances are almost always shorter than the specified distances.
Common DRI values for cameras and drones
DRI values can vary greatly depending on the specific camera and drone system. However, common benchmarks for high-performance drones (EO sensors) are usually in the following ranges:
Detection range: up to 10 km for vehicle-sized objects
Recognition range: up to 5 km for vehicle-sized objects
Recognition: up to 2 km for vehicle-sized objects
These values are examples only and may vary depending on the specific system and conditions.
Do Autel thermal imaging drones have DRI?
Autel thermal imaging drones include the Autel 640T series, the EVO MAX series, and the unreleased Autel Alpha and Autel Titan drones.
The Autel 640T series thermal imaging drones have a thermal image resolution of up to 640*512 and can recognize multiple objects simultaneously.
Autel EVO MAX series thermal imaging drones have a thermal image resolution of up to 640*512, which can identify multiple objects at the same time. At the same time, due to the rangefinder on the camera drone, it can accurately identify how far away the target is.
The DRI technology of thermal imaging drones plays an integral role in improving the performance of drone object recognition systems. By utilizing DRI technology, drones can more effectively detect and classify objects in various environments, contributing to a wide range of applications from security and surveillance to agriculture and disaster response.
