The tomato farming industry needs to adopt new ideas in applying the technology for its growth monitoring and main. Machine vision and image processing techniques have become useful in the increasing need for quality inspection of fruits, particularly, tomatoes. This paper deals with the design and development of a computer-vision monitoring system to assess the growth of tomato plants in a chamber by detecting the presence of flowers and fruits. The system also provides maturity grading for the tomato fruit. Two pre-trained deep transfer learning models were used in the study for the detection of flowers and fruits, namely, the Regional-based Convolutional Neural Network (R-CNN) and the Single Shot Detector (SDD). Maturity classification of tomato fruits are implemented using the Artificial Neural Network (ANN), K-Nearest Neighbors (KNN), and the Support Vector Machine (SVM). Evaluation results show that for the detection of flowers and fruits, the overall accuracy of the R-CNN is 1.67% for flower detection and 19.48% for the fruit detection while SSD registered 100% and 95.99% for flower and fruit detection respectively. In the machine learning for maturity grading, SVM produced the training-testing accuracy rate of 97.78%-99.81%, KNN with 93.78%-99.32%, and ANN with 91.33%-99.32%.

No abstract

The size of tomato fruits is closely related to the market segment and price. Manual sorting in tomato is very dependent on human interpretation and thus, very prone to error. The study presents thresholding, machine learning and deep learning techniques in classifying the tomato as small, medium and large based from a single tomato fruit image implemented using Open CV libraries and Python programming. Tomato images with different sizes are gathered where features like area, perimeter and enclosed circle radius are extracted. The experiment shows that using thresholding, a classification accuracy of 85.83%, 65.83% and 80% was achieved for area, perimeter and enclosed circle radius, respectively. For machine learning, the training accuracy rates were recorded as 94.00%-95.00% for SVM, 97.50-92.50% for KNN and 90.33-92.50% for ANN. Comparison of models revealed that SVM is the most model without over fitting. The deep learning approach, regardless of the algorithm, produced low performances with 82.31%-78.21%-55.97% trainingvalidation-testing accuracy for VGG16, 48.17%-41.44%-37.64% for InceptionV3 and 56.05%-44.96%-22.78% for ResNet50 models. Comparative analysis showed that machine learning technique bested the performance of the thresholding and deep learning techniques in classifying the tomato fruit size in terms of accuracy performance.

No abstract

No abstract