Open AccessColour Characterization and Detection of Dry Chinese Sausage Casing Twist using Colour Image Analysis
Author(s) -
Chee Kiang Lam,
Phaklen Ehkan,
Suwimol Jungjit
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1755/1/012050
Subject(s) - hue , artificial intelligence , rgb color model , computer vision , casing , computer science , mathematics , pixel , engineering , petroleum engineering
Chinese sausage is traditional air-dry sausage that loved by the Chinese community around the globe. This sausage can be served alone after steam cook or cook with other ingredients to create other tasty Chinese delicacies. However, a manual hand cutting process is required to cut out the sausage linking twist during the sausage’s packaging phase. This hand cutting process is tedious, time consuming and danger to workers. This study has proposed two types of sausage casing twist detector using image processing technique based on HSL and RGB colour characteristic and a set of blob detection technique to detect sausage casing twist blob on the output image. Hue, green-blue and red-blue colour characteristics are found being significant to represent the sausage casing twist in the sample images and used in the proposed algorithms. The RGB based algorithm is capable to produce a low noise image with SNR of 3.73dB as compared to Hue-based detector at -8.89dB which unsuccessful to remove shadow and object outlining of the output image. The proposed blob detection algorithm is able to detect 73.02% of the blob in hue-based image while 94.18% in RGB based image. The false detection rate in hue-based image has reached 333.33% compared to 12.17 % in RGB based image.