Blood Sample Microscopic Image Classification Dataset

Current scientific research in medical testing relies on manual analysis by experts, which is time-consuming and prone to human error. Existing automated analysis systems lack precision when processing different types of blood samples. This dataset aims to establish an efficient deep learning model to enhance the classification and detection of different types of blood cells under the microscope, meeting the needs of rapid medical diagnosis and research. Data collection is conducted using high-resolution microscopes in a standardized laboratory environment to ensure consistency in sample collection. Quality control involves multiple rounds of annotation and consistency checks, ensuring each image is accurately annotated and reviewed by experts. The annotation team consists of professionals with a biomedical background, totaling 20 people. Data preprocessing includes techniques such as image denoising, normalization, and augmentation, stored in JPG format, organized by sample category.