Fruit Recognition and Weight Scale Estimation Based on Visual Sensing


  • Budi Sugandi Politeknik Negeri Batam
  • Ria Wahyu Politeknik Negeri Batam
  • Sindi Apriliana Politeknik Negeri Batam
  • Fitri Ramadani Putri Politeknik Negeri Batam


fruit recognition, weight scale estimation, visual sensing, rgb histogram


This paper aims to develop a system to recognize fruit and estimate its weight scale based on visual

sensing. The images of fruit are captured by camera and processed by image processing to be recognized and

estimated their weight. The fruit recognition is performed based on average of RGB histogram. The RGB

histogram of each fruit is calculated and saved as training data. To evaluate the recognition process, the testing

data is compared with training data. The weight scale estmation is performed by calculating the height dan

width of the detected fruit image. The regression equation is used to determine the weight of the fruit. The

experiment was performed to 8 types of fuit with 10 samples data of each. The experiment results show the

effectivenes of the algorithm to recognize and estimate the weight of fuit with average error 9.38 % of

recognition and 4.85% of weight estimation.