Machine vision-based software for automating the grading process of philippine table eggs

Issue Date

6-2015

Abstract

Machine vision was applied for the development of software for sizing Philippine table eggs and capable of interfacing with microcontrollers. To minimize the effect of aging on egg weight and subsequent accuracy of the software, only fresh egg samples were used for the study. The best weight estimate was obtained using the weight-to-projected area relationship above all the other parameters, with R = 0.94 and R2 = 0.88. Regression analysis of weight-to-projected area data points yielded the equation: W = -4.86 + (0.04∗A) where W is the estimated weight of the table egg sample while A is the projected area. The weight-to-projected area relationship had a sorting accuracy of 92%. Results of the test for similarity of extracted parameter showed significant uniformity of values regardless of the table egg orientation with respect to the camera. Furthermore, the TAKEX TM-524NA monochrome camera performed faster than the Samsung ES-60 camera, with average processing times of 61.95 and 62.67 milliseconds (ms) for each sample, respectively. Correspondingly, the theoretical weighing capacities of the software were computed to be 58,111 and 57,443 eggs per hour for the monochrome and color cameras, respectively.

Source or Periodical Title

Philippine Agricultural Scientist

ISSN

0031-7454

Volume

98

Issue

2

Page

148-156

Document Type

Article

Language

English

Subject

Automation, Machine vision, Microcontroller, Sorting, Table eggs, Weight

Digital Copy

yes

Link to Full Text

Share

COinS