Las cámaras trampa se usan mucho en ecología aplicada, pero casi no hay protocolos para asegurar la calidad de sus datos. Solo el 4,8% de los estudios revisados aplica control de calidad. Se propone un marco para detectar errores, clasificar archivos (con o sin IA) y aplicar controles que aseguren que los datos sean confiables y listos para análisis.
1. Camera traps are a mainstream methodology in applied ecology, but surprisingly there are no widely accepted protocols to ensure the quality of the data obtained from these devices.
2. We reviewed a sample of 147 articles from the recent camera-trapping literature and found that only 4.8% report a measure of quality control.
3. We propose a framework to process media files obtained from camera traps that minimises errors by adopting a series of systematic procedures. Before classification, the focus is on detecting camera malfunctions, correcting storage and programming errors and establishing clear exclusion criteria. Classification can follow different approaches, including single or double human-eye review, which can be supported by artificial intelligence.
4. The protocol is followed by quality control procedures that enable users to determine whether a dataset meets quality standards and is ready to be analysed, or if further revision is needed.
5. Synthesis and applications: The proposed protocol introduces quality control as a key component of camera trap data processing, thus reducing error rates and making the reporting process more transparent. These principles also apply to other methods, such as autonomous sound-recording units. We suggest that by adopting formal quality control procedures, applied ecology will be able to capitalise the many advantages brought by new technologies and data processing tools.
