Oiler gesture recognition method based on pedestrian arm angle analysis

Abstract

The invention discloses an oiler gesture recognition method based on pedestrian arm angle analysis, and the method comprises the steps: enabling a monitoring system to collect a real-time video of anarea nearby an oiler in real time, and capturing an image every preset time, and obtaining a real-time image set; detecting pedestrians in the real-time image set by using a YOLO V3 algorithm, and extracting pedestrians in the images to obtain a pedestrian image set; performing category definition on the pedestrian according to the arm angle of the pedestrian in the image, 0 representing the pedestrian with the arm angle of 0-60 degrees, 1 representing the pedestrian with the arm angle of 61-100 degrees, and 2 representing the pedestrian with the arm angle of 101-180 degrees; constructing a convolutional neural network, and training the pedestrian image set by using the convolutional neural network; judging the pedestrian image set by using the trained convolutional neural network, and ifthe output is 0, judging that the pedestrian has no posture; if the output is 1, judging that the pedestrian keeps the vehicle guiding posture; and if the output is 2, judging that the pedestrian keeps pointing to the display screen.

Description

technical field [0001] The invention relates to the field of image technology, in particular to a method for gesture recognition of refueling personnel based on the analysis of pedestrian arm angles. Background technique [0002] At the beginning of the establishment of the gas station, according to the security requirements, cameras were installed in the tanker area, and the refueling operations of the gas station personnel were inspected through camera monitoring. Among them, whether the refueling staff has standard gestures during the refueling process seriously affects the image of the gas station in the minds of customers. When oil companies manage the gas station sites within their jurisdiction, they can only monitor whether the refueling staff at each station is refueling through spot checks. There are standard gestures during the process. [0003] Existing technology is to adopt the method of human intervention completely, monitor whether the refueling staff has sta...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem of supervising refueling gestures only through manual intervention in the prior art, the present invention provides a method for refueling gesture recognition based on pedestrian arm angle analysis

Images