Design and Implementation of Multi-Mode Surveillance System for Several Applications

Abstract

With the rise in crimes, thefts, stealing, burglary, and offensive violations in today's technological age, it is critical that we install cameras in every conceivable public place to discover accused / criminals / fled persons. Security agents, such as cops, use public camera footage to track down the accused's movements. People are assigned to this responsibility of watching the public camera recordings. These individuals must constantly watch the detected faces and compare them to the accused faces. However, because this is a continual process, there is a high risk of manual errors when utilizing this method. The accused person's face can also be discovered using this method, but by the time the assigned person tells the nearest police station, the accused person may have left/escaped. With all of these considerations in mind, it is necessary to avoid these blunders. Manual mistake in determining the perpetrator's face and relaying the information to officials swiftly are the main problems. Installing a camera with an already supplied data collection of suspected people can help cross-check the detected faces with the uploaded data set, which is the first issue. This improves the precision of the accused's detection. The second issue is telling officials about the suspected criminal or accused in a timely manner. If the installed camera has a transmitting unit, with the receiver on the official side and the transmitter on the camera end, this will work. The information is swiftly transmitted to the authorities, and there is a good possibility that the perpetrator will be apprehended. A frequency band is fundamental for any transmission to actually occur. The frequency ranges essential for the above-mentioned transferring data are 400–500 Hz and 900–1000 Hz. The LoRa module will be used to use these bandwidths.