Identification, detection and classification of multiclass heterogeneous blood cell series based on the council algorithm and aggregation of tissue descriptors

Understanding the functioning of biological cells and the differentiation of cells from each other is of great importance for disease diagnosis and treatment. According to expert doctors, if we can reveal the abnormality in the first stages of the formation of changes in blood cells, we will be able to treat it early and prevent its complications. In the proposed scheme, the image wavelet coefficients are fed to a YOLO neural network to distinguish between different blood types. In the following, convolutional neural network, golden eagle optimization method (GEO) and KNN classifier are used to create a new classifier. In parallel, the power of three famous classifiers including Decision Tree (DT), Simple Bayesian (NB) and K Nearest Neighbor (KNN) is used as collaborative classifier. The simulation results indicate that the presented model has accurately predicted the type of blood cell based on the training given to the model.