New from Bosch is the first camera based on its Inteox open camera platform; the MIC inteox 7100i. The new MIC inteox cameras share the same housing as the rest of the MIC cameras, and support machine learning, neural network based video analytics, and third-party software apps for customised applications. These cameras can be classified as “Driven by OSSA.”, the developers add; as the cameras follow the Open Security & Safety Alliance (OSSA) Technology Stack for video security devices, for connecting with the Security & Safety Things (S&ST) Application Store.
The cameras also allow for the creation of customised software apps that can be deployed on Inteox cameras. You can mix and match apps from the S&ST Application Store or develop software apps. Bosch has introduced two Inteox development kits containing either a bullet camera or mini-dome camera to complement MIC inteox cameras in pilot installations and support further app development.
MIC inteox 7100i cameras support predictive solutions with Artificial Intelligence (AI). It includes Intelligent Video Analytics, Camera Trainer based on machine learning, and video analytics based on deep neural networks (DNN). This AI enables Inteox cameras to understand what they see and add sense and structure to captured video data with metadata, the developers say. This process is a first step in converting video data into actionable insights and building predictive solutions, helping users anticipate unforeseen events and prevent them from happening. For example, in city surveillance, the MIC inteox 7100i cameras can detect abnormalities in behaviours, provide density data, and alert traffic operators to prevent unwanted situations. By adding third-party software apps to Inteox cameras, integrators can customise security systems, enabling them to go beyond their basic functionality. In turn, app developers can use the camera’s metadata, neural network based video analytics, and machine learning to introduce software to the market.
Traffic applications
The MIC inteox 7100i object classification (OC) models are made for Intelligent Transportation Systems (ITS) applications. Powered by the edge computation capabilities of Inteox, the OC models offer a Traffic Detector, a Video Analytics feature based on DNN that helps distinguish and classify automobiles in congested scenes. Possible disturbances caused by vehicle headlights or shadows are ignored, the firm adds. The actionable insights delivered by the cameras’ Traffic Detector help detect the presence and volume of vehicles to intelligently control traffic lights to keep traffic flowing when monitoring intersections.
Automatic Incident Detection (AID) can provide safety of people and vehicles in tunnels and on highways. Future OC models planned for June will expand upon these DNN-based features for various traffic applications.
MIC inteox cameras range from a 4K enhanced model featuring optical image stabilisation to a 1080p model in a base or enhanced version featuring Bosch’s starlight, that can capture colour images down to a level of 0.0069 lux as standard. The 4K model’s optically-stabilised images maintain pictures even when the camera is subject to vibration, such as on bridges. With the optional illuminator, the 4K model covers a distance of 300m (984 ft). The 1080p models feature 30x zoom and frame rate of 60 frames per second. When light levels drop to zero, the optional illuminator ensures detail over a distance of up to 550m (1,804 ft). Versions of the 4K and 1080p models include an in-window defroster, and onboard storage; for applications such as city surveillance, traffic monitoring, and perimeter security.
Inteox development kits
Early adopters can take advantage of Bosch Inteox development kits to complement the MIC inteox 7100i cameras in pilot installations. The kits contain either a Bosch fixed bullet (Dinion) or mini-dome (Flexidome) outdoor camera that will run on the S&ST operating system. Both kits support the testing and development of an overall system solution based on apps already available through the application store and self-developed apps.
