Indian Journal of Engineering & Materials Sciences

Open Access Open Access  Restricted Access Subscription Access

Colour discernment of tomatoes using machine vision system with OpenCV Python and Raspberry Pi


Affiliations
1 Department of Agricultural and Biosystems Engineering, NDSU, Fargo 58102, North Dakota, USA, United States
2 Department of Processing and Food Engineering, AEC&RI, TNAU, Kumulur 621 712, Tamil Nadu, India
3 Department of Food Processing and Engineering, AEC&RI, TNAU, Coimbatore 641 003, Tamil Nadu, India

Colour is the first quality attribute of food that consumers examine and it is an important component of food quality that influences market acceptance. Classification of fruits by visual inspection is an arduous, time-consuming process and prone to human error. The machine vision system is a distributed control system that integrates several machine vision modules with a control module and a user interface unit. This research proposes a method for recognising and sorting tomato fruits into a preferred location continuously. Before designing colour sorter the physical properties of fruits were studied. The major and minor diameter of the tomatoes ranges from 45-60mm and 35-50mm, respectively. The mean geometric diameter, sphericity and surface area were 48.64mm,0.94 and 6477.14mm2 respectively. The average length, width, thickness, bulk density and true density were 54.63mm, 48.44mm, 51.42mm, 0.6874g/cm3, 0.9852g/cm3, respectively. The colour sorter was researched, designed and created with Raspberry Pi, USB camera, servo motor and different digital as well as mechanical components. The model used for Raspberry Pi is Raspberry Pi 3 Model B+, USB camera with a video resolution of 640 x 480, 4.2-6V servo motor. Image evaluation is completed on each captured picture and Raspberry Pi will do the selection of which fruit can be sorted. Specific programming code in Python is written for this system. The developed colour sorter captures images and diverts fruits into the respective channel at the rate of 1800 fruits/h (i.e.one fruit per 2 seconds).
User
Notifications
Font Size

Abstract Views: 109




  • Colour discernment of tomatoes using machine vision system with OpenCV Python and Raspberry Pi

Abstract Views: 109  | 

Authors

A Astina Joice
Department of Agricultural and Biosystems Engineering, NDSU, Fargo 58102, North Dakota, USA, United States
P Rajkumar
Department of Processing and Food Engineering, AEC&RI, TNAU, Kumulur 621 712, Tamil Nadu, India
J Deepa
Department of Food Processing and Engineering, AEC&RI, TNAU, Coimbatore 641 003, Tamil Nadu, India
R Arulmari
Department of Processing and Food Engineering, AEC&RI, TNAU, Kumulur 621 712, Tamil Nadu, India

Abstract


Colour is the first quality attribute of food that consumers examine and it is an important component of food quality that influences market acceptance. Classification of fruits by visual inspection is an arduous, time-consuming process and prone to human error. The machine vision system is a distributed control system that integrates several machine vision modules with a control module and a user interface unit. This research proposes a method for recognising and sorting tomato fruits into a preferred location continuously. Before designing colour sorter the physical properties of fruits were studied. The major and minor diameter of the tomatoes ranges from 45-60mm and 35-50mm, respectively. The mean geometric diameter, sphericity and surface area were 48.64mm,0.94 and 6477.14mm2 respectively. The average length, width, thickness, bulk density and true density were 54.63mm, 48.44mm, 51.42mm, 0.6874g/cm3, 0.9852g/cm3, respectively. The colour sorter was researched, designed and created with Raspberry Pi, USB camera, servo motor and different digital as well as mechanical components. The model used for Raspberry Pi is Raspberry Pi 3 Model B+, USB camera with a video resolution of 640 x 480, 4.2-6V servo motor. Image evaluation is completed on each captured picture and Raspberry Pi will do the selection of which fruit can be sorted. Specific programming code in Python is written for this system. The developed colour sorter captures images and diverts fruits into the respective channel at the rate of 1800 fruits/h (i.e.one fruit per 2 seconds).