LabEquipVis: An annotated image dataset of computer laboratory equipment for object detection and smart lab automation

This article introduces LabEquipVis, a comprehensive dataset of high-resolution, annotated images of general-purpose laboratory equipment designed to facilitate machine learning and computer vision research in laboratory automation. The dataset comprises RGB images of ten common laboratory items, including AC units, chairs, CPUs, digital boards, fire extinguishers, keyboards, lights, monitors, mice, and projectors. Images were captured from multiple laboratory facilities at East West University, Dhaka, Bangladesh, using an Oppo Reno8 Pro camera. To ensure diversity and robustness, footage was recorded from four distinct angles: front, top-down, 45-degree, and side views, under consistent lighting conditions. Additionally, the dataset includes augmented versions of the images employing transformations such as rotation, brightness adjustment, and cropping, which enhance model generalization for real-world scenarios. All images were standardized to a resolution of 640 × 640 pixels in JPEG format. This dataset serves as a novel resource for advancing research in object detection, robotics, and smart laboratory systems, addressing a critical gap left by existing datasets focused primarily on the instruments that are used in the chemical laboratories not about the computer laboratory. It is expected to support applications in intelligent robotic assistants, automated equipment tracking, and resource management in modern laboratory environments.