Detecting people using CVEDIA-RT

What exactly is people detection

People detection uses computer vision to locate humans in videos/images. Earlier, low accuracy and false alarms were common issues. But, deep learning (AI) has greatly improved results, producing bounding boxes with confidence levels and labels to show person detection in images.

The value of detecting people

People detectors reduce false alarms in older computer vision systems by relying on AI instead of motion detection, which can be noisy and mistake non-human movements for people. The information from people detectors is used as building blocks in a larger computer vision system and fed into a tracking algorithm to follow a person over time, as a single picture can't determine if someone is coming or going.

Applications using people detection

People detectors can be found in almost every smart system out there. Basically any system that interacts with us, whether it's for safety, security or comfort will try to detect people.

Smart cities

Pedestrian crossings, automated security gates

Industrial safety

Forklift warning systems, wearing of personal protection equipment, robotic arms

Smart homes

Automatic AC and light control, doorbell notifications, home security

Perimeter security

Access control, intrusion alerts, people counting

Camera viewing angle

Most people detector models only work at low elevation angles, but CVEDIA AI models, trained using synthetic data, work from any angle, including top-down, offering greater flexibility in camera installation.

What makes a good people detector

A people detector has many qualities that determine how well it's going to work for your application. Some of those qualities are based on what type of data the model was trained on. Others are determined by the type of hardware or camera it will run on. Picking the best combinations is often a game of trial and error.

Book a demo now

Location & environment

AI models, including people detectors, are sensitive to the data used for training, so be cautious of models trained only on indoor or daytime data as they may not perform well in different environments. CVEDIA's models, trained using rendered data, are insensitive to location, weather, and lighting conditions, offering robust models and a better user experience.

Accuracy vs speed

The choice of AI model size depends on the AI chipset and the desired accuracy-performance trade-off for the specific application. Small models are lightweight and suitable for low-power devices but sacrifice accuracy, while larger models are slower but offer higher accuracy and are suited for powerful GPU's. Our wide range of performance settings enables you to pick the best model for your use-case.

Object size

Finding smaller individuals becomes increasingly challenging for a people detector, but it also enables the use of lower resolution cameras or cheaper lenses, making it a balancing act. CVEDIA-RT, our free software, helps you find the optimal strategy for your application by allowing you to experiment with different settings and models.

Start using CVEDIA-RT

See how CVEDIA's detection accuracy transforms your operations.

Get started