Examining microbial colonies on agar plates have been at the core of microbiology for many decades. It is usually done manually, and therefore subject to bias besides requiring a considerable amount of time and effort. In order to optimize and standardize the identification of bacterial colonies growing on agar plates, we have developed an open access tool available on GitHub: ColFeatures. The software allows automated identification of bacterial colonies, extracts key morphological data and generate labels that ensure tracking of temporal development. We included machine learning algorithms that provide sorting of environmental isolates by using cluster methodologies. Furthermore, we show how cluster performance is evaluated using index scores (Silhouette, Calinski-Harabasz, Davies-Bouldin) to ensure the outcome of colony classification. As automation becomes more prominent in microbiology, tools as ColFeatures will assist identification of bacterial colonies on agar plates, bypassing human bias and complementing sequencing or mass spectrometry information that often comes attached with a considerable price tag.