Combination of Canny Edge Detection and Blob Processing Techniques for Shrimp Larvae Counting

Abstract

This paper presents the proposed method to count the number of the shrimp larvae using computer technology based approach. The high demand of computer aids in counting Shrimp larvae are required in calculating survival rate of Shrimp larvae. Currently, the process of handling shrimp from post to adult larvae stage is still conducted with a tedious and labor-intensive task especially in calculating the number of shrimp larvae. However, the use of the manual counting approach has suffered with some drawbacks such as inaccurate, time consuming, laborious, human intervention, sequential processes and over-estimation. Since that, many researchers are interested into this the fundamental research and development challenges. In this study, we present an image processing technique and blob analysis for counting the number of shrimp larvae. The use of image processing technique helps for object counting specifically on image enhancement and image segmentation. Meanwhile, a Blob analysis algorithm is used as connected component to find and count shrimp larvae in the image through the experiment. The performance of the proposed method is evaluated by comparing the results with the manual counting system using all ninety samples dataset. All ninety samples of shrimp larvae (Penaeus vannamei) were obtained from the Freshwater Hatchery of University Malaysia Terengganu. The experimental results demonstrate that the proposed method is outperformed compared to the manual counting approach through the experiment by reducing the analysis time with minimizing the sequential process task.