Analyze and count with Deep Learning

Deep Learning is a revolutionary technique allowing in particular to train computer system with shape recognition. Then, the machine is endowed with artificial intelligence and shows the capacity to recognize objects like the human eye. PixAlarm proposes this technology to assure a wide range of scene analysis functions. Furthermore it is possible to use it on moving cameras, allowing to embark this function on vehicles.

Searching objects

Whether it is for people, vehicles or other type of objects, the object research is capable of spotting clearly what is present in the scene and then indicate their number. The video can be divided into zones of interest which will have their independent counter. This information can be afterward used in scenarios or be exported to generate statistics.

Intrusion detection It is possible to locate the presence of a person in private or forbidden places.	Count the number of people in a line It allows you to react quickly when too many people are present (open additional desk, call additionnal crew for help, etc.).	Count the number of vehicle in a parking You can film a parking and know in real time the amount of parked vehicles.

Passage counter

In this mode, by means of lines of crossing positioned on the image, Pixalarm allows to count desired objects (human beings, vehicles, etc.). The line can also be directional to count what passes in a particular direction. Pixalarm proposes a graphical interface dedicated to the passage counter which allows to visualize in real time the counter evolution.

Traffic analysis With a camera filming a road, you can know the number of vehicle passing in each direction.	Visit counter By placing a camera in the entrance of a store, you can measure the number of visitors. This principle can be widened to several cameras to measure the activity of the various paths or the sectors of your area.	Crowd counter The algorithm is capable of spotting a person even partially hidden by an other one, so the crowd counting becomes very efficient.

Specialized analyze

For particular needs, Deep-Learning network can be spread to every types of objects or shape.

Animal counter It is possible to count the number of animals in an enclosure or to make sure that none escaped.	Abandonned object detection With the detection of luggage and bags, it is possible to alerts if one of thoses remain motionless without anybody nearby.	Influx in transport By using existing cameras, Pixalarm can measure the influx in buses, trains or subways, in real time.

For your specific needs, please contact us.