<![CDATA[Stingless bees play an important role as natural pollinators in ecosystems and as producers of economically valuable products such as honey, propolis, and bee bread, which are utilized in the food and health industries. Identifying stingless bee species remains a challenge due to the many species with similar morphology, requiring more efficient and accurate methods. This study aims to develop an automatic system based on image processing technology for the identification of stingless bee species using Convolutional Neural Networks (CNN), TensorFlow, and the Single Shot MultiBox Detector (SSD) implemented with OpenCV. The test results showed that the developed system was capable of automatically detecting and classifying stingless bee species with an average accuracy of 98%, especially when the object was directly aligned with the camera. Out of 40 tested samples, 31 samples were recognized, and 9 samples were not, resulting in a success rate of 77.5%. Factors influencing detection success include the quality of training data, camera positioning, and morphological similarities between species.]]>
Copyrights © 2024
