Ear region selection method for call detection

Abstract

The invention discloses an ear area selection method for calling detection. Positioning a sensitive area by using the key points, wherein the sensitive area is divided into a left area and a right area; for the left sensitive area, setting the abscissa of the key point 5 as the right boundary, setting the abscissa of the key point 5 minus the 2/3 value of the face width as the left boundary, setting the ordinate of the key point 8 as the lower boundary, and setting the ordinate of the key point 17 as the upper boundary, thereby positioning the left sensitive area for making a call; for the right sensitive area, setting the abscissa of the key point 11 as the left boundary, setting the sum of the abscissa of the key point 11 and 2/3 of the width of the face as the right boundary, setting the ordinate of the key point 8 as the lower boundary, and setting the ordinate of the key point 26 as the upper boundary, thereby positioning the right sensitive area for making a call; substituting into an image to intercept a sensitive area; the sensitive area is selected on the basis of the face key points, stability is good, and accuracy is high.

Description

Technical field: [0001] The invention belongs to the technical field of driving assistance, and in particular relates to a method for selecting ear regions for phone call detection. Background technique: [0002] With the development of the automobile industry and the increase of my country's current per capita income, the vehicle ownership rate in my country is also getting higher and higher, and driving safety issues are becoming more and more important. Relevant data show that the behavior of the driver holding the phone while driving will seriously affect the concentration, which greatly increases the accident rate. [0003] At present, in order to deal with this problem, there are some solutions, which mainly focus on using the intersection camera to obtain video and then send it back to the background to detect whether the driver is making a phone call. However, this method cannot effectively prevent the illegal behavior of the driver in a timely manner. [0004] Usin...

AI Technical Summary

Problems solved by technology

[0005] Most of the current detection algorithms for driver’s hand-held phone behavior in the car use the face area to directly locate the sensitive area when making a call. Due to the large fluctuations in the face area, the selection of sensitive areas is prone to large deviations in practical applications, stability and accuracy lower sex

Images