A Microbiological Image Repository of Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa and Staphylococcus aureus Bacterial Colonies on Brain-Heart-Infusion Agar and Pseudomonas aeruginosa Bacterial Colonies on MacConkey Agar
Description
This dataset consists of images of four types of bacterial strains. The images of the bacterial colonies were taken under two different shooting conditions, “controlled” and “uncontrolled” as described in “steps to reproduce” section. These bacterial strains are: 1.Pseudomonas aeruginosa (P. aeruginosa) on MacConkey agar: the number of images under controlled conditions is (376) and the number of images under uncontrolled conditions is (2078). 2.Staphylococcus aureus (S. aureus) on Brain-Heart-Infusion agar: the number of images under controlled conditions is (153) and the number of images under uncontrolled conditions is (3573). 3.Pseudomonas aeruginosa (P. aeruginosa) on Brain-Heart-Infusion agar: the number of images under controlled conditions is (89) and the number of images under uncontrolled conditions is (2735). 4.Klebsiella pneumoniae (K. pneumoniae) on Brain-Heart-Infusion agar: the number of images under controlled conditions is (158) and the number of images under uncontrolled conditions is (2579). 5.Escherichia coli (E. coli) on Brain-Heart-Infusion agar: the number of images under controlled conditions is (173) and the number of images under uncontrolled conditions is (2888). IN THE REPOSITORY, YOU WILL FIND: Group 3 consists of images taken from 24 plates of P. aeruginosa on MacConkey agar under controlled and uncontrolled conditions. Group4 consists of images taken from: 25 plates of S. aureus, K. pneumoniae and E. coli on Brain-Heart-Infusion agar under controlled and uncontrolled conditions. 21 plates of P. aeruginosa on Brain-Heart-Infusion agar under controlled and uncontrolled conditions. An excel sheet detailing the numbers of images in the folders. NOTABLE FINDING: Baseline CNNs trained on this data achieved high accuracy, indicating that phone images provide sufficient discriminative signal without expert inspection. Refer to references below. HOW THIS DATASET CAN BE USED: This dataset can be utilized in any research interested in recognizing different features of bacterial colonies. The dataset can also be used to train and evaluate deep learning models for colony classification, while also supporting studies on robustness, generalization, and practical deployment, to advance computer vision and AI applications in microbiology. By combining clinically important bacteria with controlled and unconstrained imaging, the dataset offers a realistic and accessible resource for researchers interested in developing AI methods that perform reliably in laboratory and non-laboratory environments. IMPORTANT NOTE: This dataset is an extension of a previous dataset which can be found at [https://doi.org/10.17632/kx6gz3wmcf.1]. The expansion includes additional bacterial strains and culture media. IF YOU USE THIS DATASET, PLEASE REFERENCE THE FOLLOWING: DOI: https://doi.org/10.1109/ACCESS.2022.3221958 DOI: https://doi.org/10.1109/ACCESS.2025.3625648 DOI: https://doi.org/10.17632/kx6gz3wmcf.1 DOI: https://doi.org/10.17632/v54x8jdx5x.1
Files
Steps to reproduce
The dataset collected in this study consists of 14,802 digital images of colonies of four different bacterial strains: E. coli and K. pneumoniae, P. aeruginosa and S. aureus. These digital images were collected in the Institute of Health Technologies and Preventive Medicine–King Abdulaziz City for Science and Technology (KACST) in the Kingdom of Saudi Arabia. The bacterial strains were initially streaked on MacConkey and Brain-Heart Infusion agar plates as described above in the “Description” section. These plates then incubated at 37◦C for 24 hours. A single colony was taken with a 10 µl loop and restreaked on a fresh agar plate for increased purity. These fresh plates were incubated under the same conditions as the other plates in the dataset. Three different mobile phones were used to collect these photos (iPhone Xs Max, iPhone 11 Pro and iPhone 7). All phone cameras were set at 1080p resolution to obtain higher quality photos. The horizontal and vertical resolution of all images was 72 dots per inch (dpi). Different shooting settings were also used in the collecting process in ‘‘controlled’’ and ‘‘uncontrolled’’ environments to obtain a wider range of image qualities, camera poses, lighting conditions, etc. In the controlled environment, the distance between the camera and plate, camera pose, lighting conditions, and camera itself were fixed during image collection. In the uncontrolled environment, these conditions were not fixed.
Institutions
- King Abdulaziz City for Science And TechnologyAl Riyadh Province, Riyadh