Weight Measurement of Holothuria Scabra Jaeger, 1833 Utilizing the Surface Area of Digitized Image Captured under Water
| Autor: | Armin S. Coronado, Mary Jane Magno-Tan, May D. Maulani, Rustom R. Garcia |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
0106 biological sciences
Surface (mathematics) Pixel 010604 marine biology & hydrobiology Binary image Image (category theory) 04 agricultural and veterinary sciences Biology biology.organism_classification 01 natural sciences Holothuria scabra Digital image Linear regression Statistics 040102 fisheries 0401 agriculture forestry and fisheries Underwater Algorithm |
| Zdroj: | Proceedings of the Fifth International Conference on Network, Communication and Computing. |
| DOI: | 10.1145/3033288.3033290 |
| Popis: | In culturing sea cucumber (Holothuria scabra), weight is an essential parameter for selection of breeders as well as in determining the correct time of harvest. H. scabra eviscerate under stressful conditions and micturate when taken out of water, which leads to erratic weight measurement and obscure data to select individuals for harvest. This research aims to automatically compute the weight of H. scabra with a new algorithm that utilizes only the surface area, through the image captured by a regular camera while the specimen is submerged under water. Digital images of one hundred seventy-seven (177) healthy adult H. scabra were converted to binary images and used to measure the individual length, width and surface area through pixel analysis. The weight of H. scabra was computed using the equation: Weight[g]=C1+(C2*Surface Area), which is generated through linear regression wherein C1 and C2 has constant values of -51.840 and 3.717, respectively. Results showed that the surface area and weight of H. scabra under normal culture condition is highly correlated (R2=0.90). Moreover, error analysis revealed that the accuracy of the software in determining the length, width, surface area and weight of H. scabra was 94.46%, 94.16%, 94.07%, and 83.79% respectively. Analysis of Variance (ANOVA) showed that comparable data were obtained between actual measurements and software generated data for length (α0.05 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|