If the robot tracks people, what does it use to identify people and track specific people?

2. 1 Tracking technology based on video image processing

At present, the tracking technology based on video image processing is the most researched target tracking technology, and there are mainly the following methods.

2. 1. 1 method based on motion analysis

The methods based on motion analysis mainly include frame difference method and optical flow segmentation method. The method of inter-frame difference is to subtract adjacent frame images, and then threshold the obtained images to extract moving targets. Optical flow segmentation method is to detect the moving target by the different speed between the target and the background.

2. 1.2 Method based on image matching

The method based on image matching can identify the undetermined target and determine the relative position of the moving target. Correct interception probability and positioning accuracy are the main performance indexes of image matching. According to the matching principle, this method can be divided into region matching, feature matching, model matching and frequency domain matching.

The idea of region matching is to overlay a whole block of reference image with real-time image in all possible positions, and then calculate the corresponding value of an image similarity measure, and the position corresponding to its maximum similarity is the position of the target. Feature matching is to calculate the correlation degree of feature attribute vectors (point, edge, line segment, surface or local energy) after extracting features, and the peak value of correlation coefficient is the matching position. Model matching is to establish a suitable target model and select the position of the target through the matching result with the image. Frequency domain matching is to transform the video image into frequency domain, and then detect the motion of the target according to the amplitude or phase of the transform coefficient.

2.2 Tracking technology based on infrared

In addition to tracking technology based on ordinary video image processing, tracking technology based on infrared images also has many applications.

The research of physics tells us that any object in nature whose temperature is higher than absolute zero will radiate infrared rays of various wavelengths. The higher the temperature of an object, the greater the intensity of its infrared radiation. Therefore, in theory, any target can be detected by infrared detector, so the tracking technology based on infrared has the characteristics of strong anti-interference ability, long detection distance and no requirement for lighting conditions. Because of these characteristics, infrared tracking technology is more widely used in military fields, such as infrared guidance, airborne infrared search and tracking system, infrared thermal aiming and so on.

2.3 Tracking technology based on satellite positioning

The main idea is to obtain the absolute position through navigation and positioning satellites for tracking. The current navigation satellites mainly include GPS,

GLONASS and Galileo.

GPS is the most widely used navigation and positioning satellite constellation in the world, which was established by the US Department of Defense in the 1970s.

GLONASS is a second-generation military satellite navigation system independently developed and controlled by the Ministry of Defence of the former Soviet Union (now Russia). Similar to American GPS, this system also has civil windows.

Unlike GPS and GLONASS, Galileo is a civil navigation and positioning system that is not controlled and managed by the military. Galileo system is a project jointly proposed by EU and ESA.