Prediction of oyster volume and weight using machine vision

被引：9

作者：

Damar, S. ^{[1
]}

Yagiz, Y. ^{[1
]}

Balaban, Murat O. ^{[1
]}

Ural, S. ^{[2
]}

Oliveria, A.C.M. ^{[3
]}

Crapo, C.A. ^{[3
]}

机构：

[1] Department Food Science and Human Nutrition, University of Florida, Gainesville

[2] Computer Science Department, Bond Hall 312, Western Washington University, Bellingham

[3] Fishery Industrial Technological Center, University of Alaska Fairbanks, Kodiak

来源：

Journal of Aquatic Food Product Technology | 2007年 / 15卷 / 04期

关键词：

Machine vision; Oysters; View area; Volume prediction;

D O I：

10.1300/J030v15n04_02

中图分类号：

学科分类号：

摘要：

Fifty oysters each from Florida, Texas, and Alaska were used to develop methods that predict overall volume and weight, and meat volume and weight. Volumes (overall, shell, meat) were measured experimentally by the Archimedes principle. Densities were calculated. Using a machine vision system, the top- and side-view areas of whole oysters were determined. Various simple functions were fit to the view area data to predict whole oyster volume and weight, and meat volume and weight. A new cubic spline method was also developed to predict oyster volumes. The r2 values for the predicted oyster volumes were Texas = 0.92, Florida = 0.85, and Alaska = 0.64. © 2006 by The Haworth Press, Inc. All rights reserved.

引用

页码：3 / 15

页数：12