Augmented reality systems are currently present in multiple sectors, such as medicine, architecture, education, and entertainment. These systems enhance user interactivity when receiving information. This article describes the design and implementation of a robust augmented reality system for the “Gustavo Orcés” Natural History Museum. The proposed system uses a camera to detect people in the image using the Histogram of Oriented Gradients (HOG) technique and a Support Vector Machine (SVM) to determine whether the image corresponds to a person or not. To increase the robustness of the system in activating augmented reality with users, PIR sensors are integrated, and a presence sensor controlled by an Atmega164p microcontroller is designed. An augmented reality application developed in Unity 3D displays interactive 3D models of dinosaurs on a screen, which is activated upon detecting museum visitors, thus providing interactivity and education on paleontology. Tests were conducted on the system to verify its response to changes in lighting and the number of users.