Semi-supervised artificial intelligence human body detection embedded algorithm based on infrared array time sequence

Abstract

The invention discloses a semi-supervised artificial intelligence human body detection embedded algorithm based on an infrared array time sequence. The algorithm comprises the steps of reading infrared array data Z (t) = [zij (t)] obtained through the detection of an infrared array sensor every 300ms, and carrying out the Kalman filtering; obtaining an optimal solution data matrix Mt = [mij] at the moment t after filtering; calculating a variance sigma t of Mt and an average value as characteristic values of the temperature matrix; circularly calculating the previous steps, storing a time sequence with the length of r, calculating the average value of the temperature matrix, and calculating the variance of two characteristic values in the time sequence Rt with the length of r; and on the basis of finding out the background feature matrix, obtaining a data matrix Ms at the moment s, calculating and judging whether a person exists or not, and performing mutual conversion between a personexistence state and an unmanned state. The privacy problem is better avoided, the human body information in the static state is more effectively detected and obtained, the detection range is large, the degree of environmental influence on detection facilities is small, and the detection adaptive capacity is high.

Description

technical field [0001] The invention relates to a human body intelligent recognition method, in particular to a semi-supervised artificial intelligence human body detection embedded algorithm in certain public scenes where it is necessary to detect and judge whether there is a human body. Background technique [0002] Usually in some scenes like public toilets, it is inconvenient or impossible to directly detect whether there is a human body in a certain behavior, and it is hoped or necessary to be able to know the data information of the human body; for example The application needs to judge whether there are people in the monitoring pit in the public toilet, so that the personnel data information is displayed on the display screen set outside the public toilet, so as to transmit to people the usage information of the public toilet location in the public toilet management system, through the public toilet The management system monitors the toilet seat usage data of public t...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the existing human body recognition technology, which is easy to damage the user's privacy, the hardware cost is high, the human body in a static state cannot be effectively recognized, the detection range is limited, and the detection facilities are relatively restricted by the environment, etc. The present invention provides a better To avoid privacy issues, it can more effectively detect and obtain human body information in a static state, with a larger detection range, and to a lesser extent reduce the impact of detection facilities on the environment. Embedded algorithm for semi-supervised artificial intelligence human body detection based on infrared array timing

Images