Skripsi
PERANCANGAN USER INTERFACE UNTUK SISTEM MONITORING SMART CCTV MENGGUNAKAN METODE CONVOLUSION NEURAL NETWORK BERBASIS INTERNET OF THINGS.
Most CCTV systems are typically used for monitoring individuals captured on camera within a specific area under surveillance. In this study, we have developed an interface for CCTV monitoring using deep learning models. The interface is built using Flask as a framework to create a website-based platform. The system of this interface displays the streaming feed from CCTV cameras connected via RTSP, and it includes a feature for sending alerts via Telegram as part of the Internet of Things (IoT). The alert mechanism triggers when the system detects an unknown individual, sending a screenshot to a designated Telegram channel. During testing, we encountered a delay issue when connecting the CCTV feed to the web interface. As a result, we conducted detection tests under two conditions: direct streaming from the web interface using the laptop's webcam as the output and using the terminal with YOLO (You Only Look Once) for CCTV output. Additionally, we conducted tests to assess the delay when the system detected an event and sent a Telegram alert. The detection tests achieved a success rate of 73%. The average delay in sending Telegram alerts was approximately 1 minute. Usability testing was also conducted to evaluate the website's user-friendliness. The usability test received a score of 74, which falls into the 'acceptable' category, but there are areas that still need improvement
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2307006412 | T130884 | T1308842023 | Central Library (Referens) | Available |
No other version available