Search

KR-102963598-B1 - Assigned seat management system

KR102963598B1KR 102963598 B1KR102963598 B1KR 102963598B1KR-102963598-B1

Abstract

The present invention relates to an assigned seat management system. More specifically, the invention relates to an assigned seat management system that detects an audience member and their location through real-time object recognition, identifies the seat of the recognized audience member by comparing the audience member's location with a pre-designated seat location, and detects whether the viewing is free by comparing the seat location information with reserved seats, thereby providing safety of the facility and convenience in operational management. To this end, the present invention provides an assigned seat management system comprising: a video capturing unit installed in a screening room and capturing an audience member seated in a seat; an audience detection and location information extraction unit that receives image information captured by the video capturing unit, recognizes the audience member within the image information based on a pre-trained object recognition model, and extracts the location coordinates of the recognized audience member; and an audience location and seat matching unit that receives the location coordinates of the audience member derived from the object recognition model via wireless communication, activates actual seat layout information stored in a database, compares and analyzes the location coordinates of the actual seat member with the location coordinates of the audience member, and identifies the seat where the recognized audience member is sitting.

Inventors

  • 주양익
  • 김태욱
  • 김현우
  • 김민서
  • 장태혁

Assignees

  • 국립한국해양대학교산학협력단

Dates

Publication Date
20260511
Application Date
20240202

Claims (4)

  1. delete
  2. delete
  3. delete
  4. Audience detection and location information extraction unit; An audience location and seat matching unit that receives the audience location coordinates derived from an object recognition model via wireless communication, activates actual seat layout information stored in a database, compares and analyzes the actual seat location coordinates with the audience location coordinates to identify the seat where the recognized audience is sitting; A reservation status determination unit that, when the seat where the audience is seated is identified from the audience location and seat matching unit, receives information on the seat where the audience is seated and checks whether the seat information is reserved based on the reservation status information previously stored in the database; A free viewing detection unit that detects that an unreserved seat is seat information determined to be free viewing based on the result determined by the reservation status determination unit; A reconfirmation request unit that transmits seat information determined to be free viewing by the free viewing detection unit to the administrator terminal Includes, The object recognition model is a deep learning-based model for real-time object detection, designed to detect only human faces, and designed to detect the faces of people wearing masks; An object recognition model is generated by photographing people according to various shooting angles, locations, and lighting conditions, and training the captured images using a deep learning-based training module; The object recognition model generates a dataset that recognizes human objects from image information captured at a brightness similar to that of an indoor movie theater, taking into account the environmental characteristics of an actual movie theater; The audience detection and location information extraction unit detects audiences within an image captured by a video capturing unit based on a pre-trained object recognition model, and outputs the location of the detected audiences on the image as coordinate information to determine the location of the detected audiences; When a learning model detects audience faces in an image, it outputs in the format person_1(x1,y1,x2,y2), and performs data preprocessing to change it to person_1 (x1+x2)/(y1+y2)/2) to facilitate comparison with the center point of the seats; In the actual seat layout information, the location and range of each seat are pre-coordinated and stored on the image information based on the angle and position where the camera is fixedly installed; The actual seating chart information stores the position coordinates of the seat based on the point on the seat where the head or upper body is located when an audience member is seated; The audience location and seat matching unit determines the seat coordinates closest to the audience's location coordinates in the pre-prepared actual seat layout information as that person's seat; The position coordinates of the seats are represented in the format chair_1 (x1,y1,x2,y2) by distinguishing the seats within the pre-captured image information pixel by pixel, and the data is preprocessed to change them to chair_1((x1+x2)/2, (y1+y2)/2) to facilitate comparison with the audience's position coordinates; The audience location and seat matching unit determines that an audience member is seated in the corresponding seat if the value derived from the distance formula between the actual seat location coordinates and the extracted audience location coordinates matches a reference value or is within a predetermined error range from the reference value, and determines that an audience member is not seated in the corresponding seat if the value derived from the distance formula between the actual seat location coordinates and the extracted audience location coordinates is outside the predetermined error range from the reference value; The reservation status determination unit includes a function that receives all location coordinates of recognized audiences for each theater, counts the number of audiences within the theater based on the audience location coordinates, and compares this with reservation status information to analyze whether it matches the number of reserved audiences recorded in the reservation status information; The reservation status determination unit determines that the seat is reserved if the seat information where the audience is seated is detected from the previously stored reservation status information, and determines that it is an unreserved seat if it is not detected; The reconfirmation request unit transmits a reconfirmation request signal requesting verification from a manager and the corresponding seat information when it is determined that the seat occupied by an audience member is an unreserved seat; The reconfirmation request unit can transmit the first captured image information and the second captured image information to the administrator terminal only if the information processed through the audience location and seat matching unit, the reservation status determination unit, and the free viewing detection unit is all seat information determined to be free viewing; Images are captured and processed one by one every few minutes or seconds, but to increase accuracy, the processed information is bundled collectively to transmit only seat information determined to be free viewing for both the primary and secondary image information; An assigned seat management system including

Description

Assigned seat management system The present invention relates to an assigned seat management system. More specifically, the invention relates to an assigned seat management system that detects an audience member and their location through real-time object recognition, identifies the seat of the recognized audience member by comparing the audience member's location with a pre-designated seat location, and detects whether the viewing is free by comparing the seat location information with reserved seats, thereby providing safety of the facility and convenience in operational management. Since the outbreak of the coronavirus, there has been a trend of movie theaters implementing self-service admission due to workforce reductions. Self-service admission is a method in which the process of staff checking movie tickets upon entry to the theater is omitted, and the audience finds the theater and seat themselves and takes their seat. While this self-service admission system has the effect of reducing labor costs through staff reduction, there is a significant concern that the system could be abused, such as by allowing free admission, because the absence of staff at the entrance to check IDs of audience members watching age-restricted movies makes it difficult to detect minors watching such films. In addition, due to the lack of staff intervention, situations frequently occur where general audience members mistakenly locate the wrong theater and seat. Meanwhile, to prevent abuse of this self-service admission system, smart seat systems are being installed in some theaters. The smart seat system is implemented in a way that the seats in the theater lower only for reserved seats. However, since smart seat systems require replacing all existing seats and a large number of sensors, installation and maintenance costs are high, and this increase in costs may lead to an increased burden on customers. Prior Art: KR Registered Patent Publication No. 10-1976332 (Published May 7, 2019) FIG. 1 is a configuration diagram of an assigned seat management system according to a preferred embodiment of the present invention, FIG. 2 is an example screen of an assigned seat management system according to a preferred embodiment of the present invention. Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. First, it should be noted that in assigning reference numerals to the components of each drawing, the same components are given the same reference numeral whenever possible, even if they are shown in different drawings. Furthermore, in describing the present invention, if it is determined that a detailed description of related known components or functions may obscure the essence of the present invention, such detailed description is omitted. Additionally, while preferred embodiments of the present invention will be described below, the technical concept of the present invention is not limited or restricted thereto and can be modified and implemented in various ways by those skilled in the art. FIG. 1 is a configuration diagram of an assigned seat management system according to a preferred embodiment of the present invention, and FIG. 2 is an example screen of an assigned seat management system according to a preferred embodiment of the present invention. The present invention relates to a technology that recognizes an audience based on deep learning technology using video captured in real time by a video capturing device, and estimates the actual location of the audience using the coordinates of the recognized audience on the video. In addition, the seat where the recognized audience is seated is identified by comparing the estimated location with the seat location of the actual place. Referring to FIGS. 1 and 2, an assigned seat management system (100) according to a preferred embodiment of the present invention comprises a video capturing unit (10), an audience recognition and location information extraction unit (20), an audience location and seat matching unit (30), a reservation status determination unit (40), a free viewing detection unit (50), a reconfirmation request unit (60), and a database (70). The video recording unit (10) is installed inside the screening room and films the audience members sitting in the seats. Specifically, the video recording unit (10) is installed facing the seats inside the screening room and films the faces of the audience members sitting in the seats. The video recording unit (10) performs recording at intervals of minutes or seconds. It is preferable that the video recording unit (10) be installed at a height higher than a person's sitting height. The audience detection and location information extraction unit (20) receives image information captured from the video capturing unit (10), detects audiences within the image information based on a pre-trained object recognition model (22), and extracts the location coordin