Skip to Main Content
Blood cells characteristics recognition is an important issue for food nutrition. Food scientists want to know the detailed information of fish red blood cells, such as the major axis of a cell, minor axis of a cell, overlap rate of a cell. It would determine the scale of agglutination of red blood cells. That means the more fatigued a fish, the more agglutination of red blood cells there are. In the animal model, we determine if a medicine significantly outperform control groups via death rate of fishes within a certain number of days. In this study, red cells characteristics levels approach was proposed. It means that the fishes would die at a lower rate than the animal model due to red blood cells image recognition replacing death rate of fishes. Previous researches focused on the shape of red blood cells recognition, most of them did not discuss the cell characteristics, such as the major axis of a cell, minor axis of a cell, overlap rate of cell recognition and computation. Therefore, the main purpose of this research is through red blood cells characteristics recognition to try to replace the traditional animal model approach. This study proposed a new algorithm coded as software to proceed red blood cells characteristics image recognition. After validation, the proposed algorithm successfully recognized the major axis of a cell and minor axis of a cell. The percentage of Hemagglutination rate was successfully calculated. The study would be applied into distribution industries, which transport the live fishes, saving money because due to reducing the fish death rate. In future studies, if some of the assumptions, colored methods or chemical reactions are relaxed or revised, it may recognize different biological photos.