Classification of Image Blood Cancer by Using Multi-Training RNN

new method presented in this research to classify bone marrow based on features classification that extracted by human body. Towards this end, new features derived from image based on blood taken by microscope used in proposed descriptor: also human pose Human pose plays important role in extracted features then using these features as the blood cancer input with classifier. In this paper we focused on using retrieval image processing techniques for divided into several steps that include image acquisition, features extraction, and classification. The Retrievable Neural Network (RNN) was used to classify the segmented cells into either normal or abnormal classes based on the features selected by the genetic algorithm (GA). As a result, the classification of cells achieved an accuracy of 98.4%. Subsequently, after the manual review of blood smears, the model will act as a second reader, and it would increase the diagnostic accuracy.