A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Supervised Learning Aided Multiple Feature Analysis for Freshness Class Detection of Indian Gooseberry (Phyllanthus emblica)
A supervised learning-based simple three class freshness detection algorithm is presented in this paper for prediction of freshness of Amla samples. Six major features from the red–green–blue (RGB) and hue–saturation–vital component (HSV) colourspace and ten other minor features have been studied here with the proposed artificial neural network (ANN) model. The proposed freshness classifier is computationally light due to the use of only ANN as the major classifying tool. More importantly, the analysis is based on images captured on smart phone only, which enables portability and hence, wide acceptability of the scheme. Accuracy of classification higher than 96.5% is achieved using the hue histogram of the image, followed by green layer histogram and others. All the major features were able to produce more than 83% efficiency in freshness class determination; whereas, minor features could achieve a highest classification accuracy of about 77%; clearly suggesting advantage of the major set of features. High efficient freshness categorization, combined with ease of computation, simple feature analysis and use of smart phones for image acquisition ensures its high possibility of real life implementation, especially incorporating within mobile application-based software development.
Supervised Learning Aided Multiple Feature Analysis for Freshness Class Detection of Indian Gooseberry (Phyllanthus emblica)
A supervised learning-based simple three class freshness detection algorithm is presented in this paper for prediction of freshness of Amla samples. Six major features from the red–green–blue (RGB) and hue–saturation–vital component (HSV) colourspace and ten other minor features have been studied here with the proposed artificial neural network (ANN) model. The proposed freshness classifier is computationally light due to the use of only ANN as the major classifying tool. More importantly, the analysis is based on images captured on smart phone only, which enables portability and hence, wide acceptability of the scheme. Accuracy of classification higher than 96.5% is achieved using the hue histogram of the image, followed by green layer histogram and others. All the major features were able to produce more than 83% efficiency in freshness class determination; whereas, minor features could achieve a highest classification accuracy of about 77%; clearly suggesting advantage of the major set of features. High efficient freshness categorization, combined with ease of computation, simple feature analysis and use of smart phones for image acquisition ensures its high possibility of real life implementation, especially incorporating within mobile application-based software development.
Supervised Learning Aided Multiple Feature Analysis for Freshness Class Detection of Indian Gooseberry (Phyllanthus emblica)
J. Inst. Eng. India Ser. A
Sarkar, Tanmay (author) / Mukherjee, Alok (author) / Chatterjee, Kingshuk (author)
Journal of The Institution of Engineers (India): Series A ; 103 ; 247-261
2022-03-01
15 pages
Article (Journal)
Electronic Resource
English
Correlation-Aided 3D Vector Distance Estimation-Based Quality Assessment of Indian Gooseberry
| Springer Verlag | 2022
Effect of phyllanthus emblica biodiesel based lubricant on cylinder liner and piston ring
| British Library Online Contents | 2018