Development of “Smart Eye” – Smartphone Application – To Determine Image Color and Texture of Tomatoes

Abstract

In this research, an image processing program based on Android application ‘SmartEYE’ used to measure image color and texture was developed. Tomatoes in three maturity classes (turning, light red, and red) were used as samples. Image processing program was developed using Android Studio in Java language program. A mobile phone captured images in 8 bit color format saved in jpeg (joint photographic experts group) format. Image features were then extracted including Lab color and texture features such as entropy, energy, contrast, and homogenity. The Lab color channel and texture features were obtained using OpenCV library function and Grey Level Co-occurrence Matrix. Results shows that Lab and chroma (C) values increase as the maturity class increase. Tomatoes in three different classes have different image textures, especially for entropy and contrast, while for homogenity values there are no significant different among the three classes. Using the developed program, tomatoes can be classified based on a and b values.