Abstract
During the critical situation of a patient, an immediate blood transfusion from others to a patient is necessary. The blood groups are AB, B, A and O. The identification of the blood group and Hb- anemic help to decide blood transfusion during emergencies. In order to check anaemia as well as to take complete blood count and to monitor a number of diseases, the hemoglobin test is an important one. But both tests are not possible in remote and village areas, because they require additional equipments and trained personnel which can be afforded only in lab. The conventional test methods involve the risk of infection, possibility of occurring human errors, more time consumption and noise interference. By addressing these limitations, an effective hardware system is proposed to detect blood group and Hb-anemic by exploiting image processing mechanism with FPGA based VLSI process. The front-end model was developed using MATLAB software, where the input blood sample images were converted into gray-scale and perform GLCM model to generate the statistical parameters. The back-end model was developed using XILINX FPGA Spartan 6 kit/ Altera Quartus Ⅱ. The data with code is then fed to hardware kit to compare and correlate with the training samples using Stump Boost Regression Algorithm (SBRA). We considered World Health Organization (WHO)’s HCS method to detect Hb-anemic. The proposed system shows improved performance, optimal accuracy, less error rate and less time with low cost. This system can be used as a technician-free self-testing device, minimising the exposure of patients to clinical settings.
Keywords: Anemia, Blood Classification, Hemoglobin Color Scale, Stump Boosted Regression Algorithm.