This study presents a novel approach for classifying white blood cells (WBCs) using extensive pre-processing with data augmentation techniques and an optimized convolutional neural network (CNN) model. The research aims to improve the accuracy and efficiency of WBC classification in microscopic images. The study compares the performance of the proposed CNN model with state-of-the-art transfer learning models such as VGG16, InceptionV3, MobileNetV2, and ResNet50. The results show that the proposed CNN model achieves a significantly higher accuracy of 0.9986 compared to the transfer learning models, which have lower accuracy scores. The proposed model also demonstrates superior performance in terms of precision, recall, and F1 score. The study also addresses the issue of overfitting by employing data augmentation techniques to balance the number of samples for various classes. Additionally, the study uses a customized CNN model with different filter sizes and layers to optimize the classification process. The proposed model is tested on a dataset consisting of five WBC classes: monocyte, neutrophil, eosinophil, lymphocyte, and basophil. The results show that the proposed model outperforms other models in terms of accuracy, precision, recall, and F1 score. The study also conducts a comparative analysis of the computational complexity of the models, showing that the proposed CNN model has the lowest training and testing time. The study also performs an ablation analysis to evaluate the contribution of the pre-processing component to the model's performance. The results indicate that pre-processing plays an important role in improving the accuracy of all learning models. The study also validates the effectiveness of the proposed approach using an additional dataset for malaria parasite screening, achieving an accuracy score of 0.9996. The proposed model demonstrates superior accuracy and computational efficiency compared to existing state-of-the-art works. The study concludes that the proposed CNN model is a practical alternative to WBC classification methodologies, emphasizing efficiency without compromising essential diagnostic accuracy.This study presents a novel approach for classifying white blood cells (WBCs) using extensive pre-processing with data augmentation techniques and an optimized convolutional neural network (CNN) model. The research aims to improve the accuracy and efficiency of WBC classification in microscopic images. The study compares the performance of the proposed CNN model with state-of-the-art transfer learning models such as VGG16, InceptionV3, MobileNetV2, and ResNet50. The results show that the proposed CNN model achieves a significantly higher accuracy of 0.9986 compared to the transfer learning models, which have lower accuracy scores. The proposed model also demonstrates superior performance in terms of precision, recall, and F1 score. The study also addresses the issue of overfitting by employing data augmentation techniques to balance the number of samples for various classes. Additionally, the study uses a customized CNN model with different filter sizes and layers to optimize the classification process. The proposed model is tested on a dataset consisting of five WBC classes: monocyte, neutrophil, eosinophil, lymphocyte, and basophil. The results show that the proposed model outperforms other models in terms of accuracy, precision, recall, and F1 score. The study also conducts a comparative analysis of the computational complexity of the models, showing that the proposed CNN model has the lowest training and testing time. The study also performs an ablation analysis to evaluate the contribution of the pre-processing component to the model's performance. The results indicate that pre-processing plays an important role in improving the accuracy of all learning models. The study also validates the effectiveness of the proposed approach using an additional dataset for malaria parasite screening, achieving an accuracy score of 0.9996. The proposed model demonstrates superior accuracy and computational efficiency compared to existing state-of-the-art works. The study concludes that the proposed CNN model is a practical alternative to WBC classification methodologies, emphasizing efficiency without compromising essential diagnostic accuracy.