Method for automatically detecting range of camera and table lamp using method

Abstract

The invention discloses a method for automatically detecting the range of a camera. The method comprises the following steps: S1, collecting an image; S2, extracting human body posture data from the acquired image by using a human body posture recognition algorithm, and if the extraction succeeds, performing a step S4; if extraction fails, executing the step S3; S3, sending out a prompt to remindthe user to sit upright, and executing the step S1 again; and S4, detecting the positions of the two shoulders, the horizontal included angle and the neck in the image according to the human body posture data extracted in the step S2 so as to determine the position of the camera relative to the user. The invention further discloses a table lamp using the method, the parallax problem of the monocular camera is eliminated, the difficulty of determining the position relation between the camera and the user is solved, and the user is reminded to correct the posture. The distance parameter betweenthe camera and the user is preliminarily determined by detecting the skull size of the user, so that the approximate position shown by the eyes of the user is found, and the user can read and follow light by matching with the table lamp body provided with the adjusting part.

Description

technical field [0001] The invention relates to the technical field of smart appliances, in particular to a method for automatically detecting the range of a camera and a desk lamp using the method. Background technique [0002] In recent years, parents have paid more and more attention to their children's education, and the time for teenagers to study has also increased relatively. Improper sitting posture will lead to problems such as myopia and hunchback of young people. At present, there are sitting posture recognition desk lamps based on the human body posture of the monocular camera in the market. The human body sitting posture recognition algorithm of the monocular camera can refer to "Discrimination Based Research on Sitting Posture Visual Recognition Method Based on Deep Learning", its classification number: TP391.41; TP18; however, it is difficult to grasp the position of the camera relative to the person in the application of the monocular camera, so most desk lam...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the following problems in the prior art: 1. It is difficult to judge the relative position of the camera and the user due to the parallax of the monocular camera; 2. Although the desk lamp of the binocular camera can avoid the parallax, it is expensive and inconvenient to adjust. The present invention designs a method for automatically detecting the range of the camera based on a monocular camera and uses the desk lamp to detect the relative position of the user and the desk lamp by using the monocular camera, and realizes the pre-information collection and the function of correcting the user's sitting posture through the detection results. User read light tracking function

Images