An artificial neural network (ANN) model for the cell density measurement of spirulina (A. platensis)

• Main Authors: Aquino, Aaron U., Fernandez, Matthew Edward M., Guzman, Aileen P., Matias, Albert A., Valenzuela, Ira C., Dadios, Elmer P.
• Format: text
• Published: Animo Repository 2019
• Subjects: Freshwater algae; Cytometry; Neural networks (Computer science); Manufacturing
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1896
• Institution: De La Salle University

This paper presents a dynamic model for the cell density measurement of Spirulina platensis by using backpropagation-based Artificial Neural Network (ANN). A vision system, composed of a camera and a photodetector, is developed to measure the color features and illuminance of the algal culture, which will then serve as the training data. The input parameters are the RGB values and the lux value from the vision system. The network has three layers with structure 4 - X - 1, where the node size X of the hidden layer is varied experimentally. After several trials of training, the model with 24 nodes showed the lowest mean squared error of 0.0047813 and fastest learning time of 2 seconds. This model was validated by performing F-test on the actual dataset and the output from the model. Results show that there is no significant statistical difference between the two, and that the output from the ANN is valid. © 2018 IEEE.