Deep learning-based classification of microalgae using light and scanning electron microscopy images.

Water pollution is one of the biggest problems at the present time. To determine this pollution, it is used classic methods generally and there are different tables which contain lower and upper limits. The results hold equal significance even if they are close or far away while obtained results are evaluated according to these tables. Thus every quality parameter located in one of the classes in the table. In this research, water samples were taken from Beykavagi Pond on a monthly basis between the dates April 2010-Mart 2011. It was measured 27 physico-chemical parameters totally as Temperature, Dissolved Oxygen, Turbidity (NTU), Electrical Conductivity, Biological Oxygen Demand, Chlorophyll-a, Ammonium, Total Phosphate, Chloride,  Magnesium, Calcium, Nitrate, Nitrite, Potassium, Sulphate, Total Hardness, Aluminum, Copper, Zinc, Iron, Cadmium, Cobalt, Chrome, Lead, Manganese  and Nickel. According to results obtained in Beykavagi Pond, index model was used in which water pollution control is evaluated with fuzzy logic as well as classic method and water quality classification. As the result of analysis made with fuzzy logic, water quality divided into four groups (I, II, III, IV). Consequently due to fuzzy logic statement, in Beykavagi Pond was provided the homogeneity between measured parameters and it was seen that 66% of quality index is in II category, 34% in III category when we calculated annual quality index which we obtain according to output value. Also in evaluation of water quality parameters as a whole it's thought that fuzzy logic method can be used as a tool. With this research, it was presented first findings for creation of fuzzy logic based index for water quality in our country.

N.Emre Gumus

Papers

Deep learning-based classification of microalgae using light and scanning electron microscopy images.

Evaluation of Water Quality Using Fuzzy Logic in Beykavagi Pond, Turkey