Overview
Images and videos are a key to most workflows in law enforcement and surveillance. These assets are being collected from a growing number of sources, including, social media, security cameras, unmanned aerial vehicles (UAV), satellite imagery, body-worn cameras, traffic cameras, etc. In parallel, complexity has also increased. It used to be enough to keep an eye on a camera and see what is transpiring at the moment. Today, the threat landscape is much more sophisticated, and because of that one needs to also have a good grasp of what may have transpired in the past, so that combined one can asses and be ready for what may happen in the near future. In other words, the goal-post has moved from "reacting" to a situation to that of "preventing" a situation.
piXserve can help your organization make sense of the data collected, and improve staff productivity. The ability to search, discover, and correlate visual information "on the fly", is now at your fingertips.
piXserve can help you make the most of your archives and live feeds and help you discover content and patterns "on-the-fly". The software provides functionality that is key to Law Enforcement organizations, including:
piXserve includes state-of-the-art facial recognition technology that allows users to search for faces and automatically name faces in images and videos. For example, video feeds from street level security cameras can be used to identify specific individuals at designated locations. A name can be assigned to a person so that instances of the individual can be found not only by using an example picture of the person, but also by typing their name in a search box. Cross correlation between pictures and videos is supported (for example, an image from a social media site can be used to search video and vice versa).
piXserve includes state-of-the-art text recognition technology that allows users to search for text strings that may be visible in the field of view of the image or video. Text in a broad range of languages can be searched using piXserve. Use this capability to search for license plates, street signs, name badges, text on a t-shirt, scrolling text from a news video feed, a picture of a tweet, ...anything.
piXserve extends the state-of-the-art in object detection and recognition. The software contains a growing pre-packaged set of automatically detected object categories. With "drag and drop" simplicity. customers can also deploy their own deep learning models, trained on their own data. and piXserve will use them as part of its image analysis pipeline. The rich set of metadata that piXserve creates allows customers to search with a level of specificity that is not possible when using classifiers alone.