Qualification Examination of Larvae Crab by Motion Detection Process

Abstract

This paper describes motion detection algorithm and technique to examine the quality of mangrove crab larvae by detecting its movement quality using image processing method.  OpenCV program was used to examine the quality of the larvae and mobile phone with 20x magnifying lens was used to record its movement during Zoea 1 (first three days after birth) with size range of approximate 40-50 microns (µm). The recording video files were processed with Object Detection, Object Filtering, and Object Tracking, respectively. 282 of mangrove crab larvae from five breeders were used in the study. Each breeder was cultured in separate truck. The larvae were randomly selected 10 times from each truck - 50 times in total. Their movement were then detected and tracked to examine the quality by calculating mean score of the movement. It was found that this method can be employed to detect and track moving objects well.

The result from examining the movement quality revealed that the larvae from the third truck showed the highest moving rate with 87.92%. It was likely that the truck contained the larva with the best developmental quality. It was then followed by the fourth truck with 40.44%, the first truck with 27.66%, the fifth truck with 23.82%, and the second truck with 19.51%, respectively.