By Vaishnavi Vijay
Visual Recognition (Computer Vision/ Visual AI) allows users (computers/ Machines) to acknowledge, grasp and analyze the contents in digital images, videos. by answering the most important questions relating to images, and videos such as ‘What?’, ‘Where?’, ‘When?’, and ‘How?’. it seeks to understand and automate tasks that the manual human visual system can do. Just directly click an image and submit it. After that, our visual recognition algorithm- Gryphos (a visual search platform), offers an unblemished experience to users by delivering precise results in real time.
Using Prisma’s visual search platform, solutions used for individuals who have issues like walking bent body or other health issues like Camptocormia Condition are being detected frame by frame using live video and by using body pose estimation (BPE) that marks the key body joints of a person. This technology smartly evaluates if any person’s upper body is bent and is walking slowly or potentially has camptocormia. Later the officials check and confirm, sending an alert to a customized mobile app with necessary respective information. In the same way if any person is finding it difficult to walk or limping, using the smart system (Artificial Intelligence), it will detect(predicts) the person’s walking pattern is abnormal and they are struggling to walk normally. After detection is done, The confirmation will be sent through an alert on a customized mobile app with the respective information.
At the same time, using Gryphos system, if any person is found walking using any aid devices like walker, crutches etc. will be detected by Gryphos’ Object detection system. After aid devices are detected, system will send an alert through the mobile app to employees or personnel assigned to assist individuals, after which the employees will be assigned to provide assistance to the person who need help. Gryphos’ system also detects all humans carrying an infant. The AI system smartly estimates if the person is carrying an infant in his/her arms. It also applies for pregnant women and people who fall by tripping down will be detected by using the same system.
The development for floor assistance is also known as ‘Desk of Goodness’ and with the help of surveillance camera detects the passengers like Senior citizen, Disabled, Pregnant Women, etc. which will help them if they are in need of anything.
The process is used through mobile applications and certain processes are being followed for airport employee staff, that starts like an airport employee should select and download the application. Next the employee needs to register themselves in the application for the first time and fill in the required credentials. Then, The employee needs to add the login credentials in the login page and once the credentials is verified in the database, if incorrect, the staff member will be redirected to the registration page and If matched then they will be directed to Homepage of the application.
The Live Cameras will be monitoring and detecting whether senior citizens, pregnant ladies, women carrying newborn babies, or disabled individuals are in the premises. If any such person is detected, then a new assistance ticket is formed within 60 seconds of countdown timing. If the 60 seconds time limit has exceeded, an alert will be send to the supervisor in the form of vibration and will notify in a red color. Then, it’s the supervisor’s responsibility to assign the case.
The staff member will be given the list of assistance to be accomplished. They need to select one ticket and choose whether to accept or decline the assistance, if he/she selected “decline” assistance the process will be ended from his end and he needs to specify the reason. If the passenger chooses accept the assistance, the staff member will be provided with the passenger’s information, where the staff member will ask for assistance required. If the passenger declines the help, the staff member will click the button saying passenger declined and the process will be ended. If the passenger says yes for help, the staff member need to select assistance in progress button. And provide the respective help. Once the assistance is completed, the staff member needs to click the assistance completed button while specifying the feedback. The staff member can view their history, check their assignments of the day, date and time.