What Is Object Detection?

Object Detection Explainer

Object detection is a computer-vision task that identifies which objects appear in an image and where they are located. A typical result contains a class label, a confidence score, and a bounding box for each detected instance.

Two-stage detectors first generate candidate regions and then classify and refine them. One-stage detectors predict classes and locations more directly. Transformer detectors instead learn a set of object queries and match them to labeled objects during training.

Detection supports traffic analysis, robotics, retail inventory, wildlife monitoring, manufacturing inspection, and assistive technology. It offers object-level locations without the expense of labeling every pixel.

Small, crowded, partly hidden, or unfamiliar objects remain challenging. Thresholds and duplicate-removal rules affect results, and training data can omit important environments. Evaluation should examine precision, recall, localization, class balance, and failures under changed conditions.

Object detection finds and classifies multiple objects in an image, usually by predicting a box and confidence score for each one.

Key terms include bounding box, class label, confidence score, region proposal, one-stage detector, non-maximum suppression, precision and recall.

The model must name each visible object and estimate its position.

Confidence thresholds and overlap handling turn raw predictions into a final set of detections.

Occlusion and small scale can make separate objects look like one region.

Per-class recall and localization errors reveal problems hidden by an overall score.

Explore more "Explainers"

Discover additional explainers across politics, science, business, technology, and other fields. Each explainer breaks down a complex idea into clear, everyday language—helping you better understand how major concepts, systems, and debates shape the world around us.