An automatic counting system for transparent pelagic fish eggs based on computer vision

• We present an automatic method for segmenting and counting live transparent pelagic fish eggs.
• An integrated morphological method for recognizing eggs is proposed.
• We improve the watershed algorithm to solve the problem of over-segmentation.
• By ImageJ tool, we count the residual eggs for manual correction.
• A comparison between the automatic and manual counting is performed.

Alive eggs of marine species with pelagic eggs float, while dead eggs usually sink. A non-invasive method for counting the number of floating eggs therefore gives the possibility to track survival throughout experiments. In this paper we present an automatic image analysis method for counting live pelagic eggs of marine fish. Pelagic fish eggs are typically transparent and difficult to detect in images. Current image analysis methods for counting pelagic fish eggs are therefore done on eggs transfixed in a polymer to create contrast between the eggs and the background. This kills the eggs. The main advantage of the presented method is that it is non-invasive and only requires a minimum of handling of the eggs. As case studies we collected images of Atlantic haddock (Melanogrammus aeglefinus) and Atlantic cod (Gadus morhua) eggs. The eggs in the images were manually counted for verification of the methodology. The average counting error of false positives was 6% and the average counting error of false negatives was 2%. This demonstrates that the method is objective and accurate.