Automatic Fish Egg Counting System Using Image Processing Techniques

Counting the quantity of fish eggs is essential in the fish industry and research sectors for determining a variety of reproductive factors such as survival rate, reproduction rate, species of fish, and so on. However, manual counting of the quantity of fish eggs is a time-consuming and challenging task. This paper presents an automated system for counting fish eggs using image processing techniques. A step-by-step procedure is outlined, involving preprocessing techniques like contrast enhancement, morphological operations, gamma enhancement, etc., followed by egg segmentation using adaptive thresholding and different morphological operations for segmentation. After that eggs are identified. In the present work, the method is applied for counting eggs of Rohu (Labeo rohita). For this, we have developed a dataset of 17 fish egg sample images. Each sample of egg is placed into a transparent dish with 10 mm depth of water. We have dispersed eggs in the dish. which are captured from the top side of the dish using a smartphone camera. The efficiency of the automated approach compared to manual counting methods. The system’s performance is evaluated through experimental analysis, demonstrating faster processing times and reduced human effort.