It is a software technology installed in the security CCTV that judges the phenomenon based on the time and the sound source. If the existing CCTV provided only the visual part, the sound source can determine the risk factors that are not in sight. IVS's on-device AI product (ISD-1000iOK-C), which integrates an image analysis camera and sound source collection device, detects abnormal sound sources such as screams, vehicle crashes, and rescue calls. It can be said that the blind spot of the existing CCTV, which relied only on visual functions, was supplemented.

This is combined with artificial intelligence to automatically detect dangerous situations. In the existing CCTV, it was difficult to respond immediately if security personnel did not check the control screen in every situation. The AI ​​CCTV can immediately recognize the occurrence of a specific situation and notify the security system. Security personnel can only respond to specific situations, so security work efficiency is high.

Artificial intelligence CCTV does not end with phenomenon detection. Continuously accumulated CCTV data greatly improves performance through deep learning technology. Therefore, prediction function based on big data is possible. After detecting the risk-generating factor, a pre-prediction of the occurrence of a situation that may occur as a result is made. Preemptive prevention rather than response after an incident has become possible. IVS expects its sales to exceed 10 billion won this year from 8.7 billion won last year amid the industrial crisis caused by the prolonged corona virus. It is also the reason why it was selected for the global ICT future unicorn fostering project.

The demand for CCTV is high in the defense sector, where the security system is the most important. The country where CCTV is applied the most is in defense after transportation. However, as the number of soldiers stationed at all times is gradually decreasing, the demand for security devices required at the Armistice Line is bound to increase. Therefore, artificial intelligence CCTV to replace manpower in national defense is essential. Since last year, the Defense Acquisition Program Administration is confident that it will be possible to raise the level of state-of-the-art defense security as the AI ​​CCTV technology has been adopted in the rapid trial acquisition project to apply civilian new technologies to national defense.

The artificial intelligence CCTV, which was ordered as a pilot project this time, will prepare a new bridgehead for national defense security through the convergence of video and sound solutions. The four image solutions to be used here are an intelligent image analysis solution (IVS-1000), an intelligent automatic tracking solution (IVS-1000AT), an intelligent image restoration solution (IVR-1000), and an intelligent image counting solution (IVS-1000CT). It has the characteristic that it is not affected at all from the surrounding detection obstacles as well as the accurate detection of the object.

Considering the fact that the existing CCTV had a false recognition rate for enemy intrusion or roaming, we supplemented the detection error through deep learning technology. In addition, with video analysis technology tailored to the defense field, the movement direction of a specific object is tracked to automatically track the enemy's intrusion. In-house image restoration is possible even in situations where object identification and image resolution may decrease due to bad weather such as heavy rain, heavy snow, or fog.

A sound function solution that eliminates the blind spots of existing CCTVs will also raise defense security. The intelligent voice and sound analysis solution (ISD-1000i), an acoustic solution, also recognizes objects that cannot be seen with the naked eye as sound sources, enabling immediate response to the enemy's camouflage intrusion. Emergency voice analysis algorithms such as ‘Help’ and ‘Save Me’ based on deep learning technology also enable prompt action due to the nature of national defense. It is expected that AI CCTV will be applied to several areas of the cease-fire line by inserting AI software into the existing military CCTV.

The AI ​​box (BOX) method that stores AI CCTV big data in a small chip type will relieve the burden of data storage. In the meantime, the increasingly massive artificial intelligence data management server occupies a fairly large space, which has become a burden both physically and financially. The simplicity of data management will promote the commercialization of big data-based CCTV. As it is a method that only needs to insert an AI box chip into the existing CCTV, there is no cost burden due to huge device replacement. Our transportation infrastructure data analysis, AI-based traffic signal control, industrial facility environment, and disaster management fields will be the cornerstones of promoting safe social infrastructure based on AI box.

Until now, the main functions of the AI ​​box have been limited to simple capture functions such as the number of people and intrusion recognition functions. As the functions that can be operated gradually become more advanced, the data management simplification function becomes very meaningful.

Personal CCTV management will be easier based on cloud services. In the past, CCTV management and maintenance required considerable costs, making it difficult for ordinary households and small business owners to perform CCTV maintenance. However, from now on, CCTV hardware, software, and database can be managed holistically in the cloud. We lowered the threshold of the personal CCTV market by developing a cloud-based intelligent video control system (IVS VSaaS-1000). Users of the cloud-based intelligent video control system can receive management tasks such as system maintenance and upgrade by paying only a monthly fee for cloud service after installing CCTV on site. Since it is possible to connect unlimited devices to the cloud, it is possible to check the CCTV video in real time anytime, anywhere through a smartphone. As physical restrictions are removed and prices are set reasonable, government ministries and corporations as well as individual consumers will be free to use AI intelligent CCTV.

As it is easy to use and apply big data, we plan to target the customized overseas CCTV market. The key is to apply our big data to CCTV functions required by country and company. There is a limit to the development of customized CCTVs with artificial intelligence alone. The local R&D center will be put together to create big data to implement customized AI CCTV.

It plans to establish a local R&D team by taking the European market centered on North America, South America and Germany as three major bases. The lack of synergy between R&D and artificial intelligence is also being pointed out as a limitation of the domestic security industry. It is an artificial intelligence CCTV technology that converges big data and has passed the first technical special evaluation, so it is preparing to go through the second round of listing. Having big data is a key factor in which competency can be expressed in preempting the global CCTV security market. Even in the non-face-to-face era, corporate sales are on the rise, which is why sales of up to 100 billion won within the next three years can be expected.