- Indonesia
Copyright © 2025 Powered by BCI Media Group Pty Ltd
Confirm Submission
Are you sure want to adding all Products to your Library?
Contact Detail
21 Jun 2021 by VIVOTEK
Traditionally, people counting at shopping malls is done manually, a responsibility assigned to security personnel or guards whose primary task is to ensure public safety and protect the property of the mall and its tenants. To carry out this task, security personnel use a pushbutton that they activate whenever a person enters the building. However they have other tasks to take care of, which prevent full reliability and effectiveness in the counting process.
Thus shopping malls are faced with a dilemma. It is important to count the number of people that visit them in order to determine the traffic flow and their potential customers, and even to determine issues such as the rent of particular high-traffic premises. Further, a number of marketing, infrastructure, events and other type of decisions are made based on the counting metrics. However, by employing security staff for this task, valuable human resources are taken up without a satisfactory end result.
This project had two stages of implementation. The first was the establishment of infrastructure: the wiring, pipes, switches and the communication elements. The second phase, nearly a week long, focused on the installation of the cameras. It was also necessary to hold another week of testing to compare the results with manual counting and video recordings.
The solution was installed in the six pedestrian entrances of the mall. Six units of VIVOTEK’s stereo network camera, the SC8131 were installed – linked intelligently together by VIVOTEK’s professional video management software, VAST.
Carlos Sanchez was extremely satisfied with the results: "Each installed camera works independently collecting its coverage statistics and reporting to the server regularly.” These individual cameras, smart on their own, become even more powerful and flexible when connected. As Sanchez added, “The reports can be exported from the server and organized in different formats.”
Sanchez, always focused on reliability, took great pains to confirm that if the camera performance parameters were met, the percentage of accuracy would be around 98%: "For example, we carried out tests with people who walked in hugging each other, and the camera detects them as two people when the heads are separated." The cameras are ideally mounted at a height of 3.6 meters and at a distance from the entrance of approximately 5 meters. When mounted in this way, the camera offers the ability to discriminate objects by their height. For example, supermarket carts are not detected as people. The hi-resolution power of VIVOTEK’s SC8131 can also differentiate shadows from people. This sets the SC8131 apart from other analytics-focused cameras which work only when the pixels in the image change; in such cases shadows interfere with metrics.
"This camera has two lenses and therefore works with a sort of triangulation, which allows for depth perception like the eyes of human beings. As shadows have no depth, the camera is not confused by them," says Sanchez.
The ICONTRONIX manager is also excited that reports generated from VIVOTEK’s system can be arranged according to the customer's needs. "We have a counter per camera, so it is possible to know the number of people going through every entrance, all entrances, or two particular entrances. A variety of options are available to suit any need.”