In intelligent video surveillance product, by processing the images taken from the cameras, the presence of any abnormal, threatening, problematic and unexpected conditions is notified to the customer. Abnormal conditions in each environment can be different, for example, a program that detects the presence of a fixed object in the unauthorized range and announces the abnormal conditions, such that the alarm of the occurrence of these conditions is announced depending on the customer's opinion. After the alarm is sent, the correctness of the incident is easily confirmed or rejected by the operator based on the review of the monitored documents, and after that the correction process starts if needed.
Artificial intelligence has given high capabilities to image-based event detection algorithms. In the product of the intelligent video surveillance system, the use of such a high potential provided by the advancement of artificial intelligence technology provides effective help to reduce human error, eliminate costs and manpower, and provide 24-hour service in all conditions for security systems.
Such systems cannot be implemented easily by foreign products due to security issues. Therefore, due to the high record that the team members have in implementing image processing and deep learning algorithms, this product brings accuracy, efficiency and high economic profit.

• In general, these functions are a subset of the following functions that are specially prepared in Roshan Tolo Sharq company in the form of modules and can be easily called in the system structure according to needs and definitions:
       1. Identification of fire - smoke and explosion
       2. Classification of door status in open and closed classes
       3. Human identification
       4. Detection of safety equipment (including hats, gloves, masks and vests)
       5. Face detection and identification
       6. Cell phone detection
       7. Detection of additional means
       8. Tracking
  
  

• 1- Designed as a web service
  2- Defining new users with different access levels
  3- The possibility of defining various event-behavior structures such as alarm-pup-up-message, ...
  4- It has a variety of selectable image processing modules based on expected functions in each environment and for each camera separately
  5- Real-time performance
  6- Bandwidth management
  7- The possibility of searching the saved images based on the type of event, time and camera number
  8- Save the report extracted from the search process
  9- Effective user interface for registering cameras, assigning desired processing modules for each camera, monitoring camera status history, announcing events, choosing the appropriate behavior for each event, reporting and archiving data.