The work in this paper defines the problem with the application of intelligent methods for combining data from sensors in the analysis of agricultural and food products. Results and comparative analyzes from the researches of the indicated products with optical, ultrasonic, capacitive sensors and sensor devices, as well as combinations of them are presented. The features according to which the classification of the studied objects according to the indicated characteristics are performed, as well as a certain set of relations between them are defined. Methods and tools for classification and prediction of main indicators for product quality have been developed. The obtained results demonstrate that the algorithms used in determining the optical, ultrasonic and dielectric characteristics in the analysis of some food and agricultural products allow the development of adequate classification models for rapid and non-destructive determination of the main quality indicators of these products. Through the obtained results, it is proved that the classification methods and the predictive models manage to keep the values of the classification error low. It was found that the informativeness of the features, the choice of methods for reducing the volume of data and ways of classification and regression, depend largely on the product under study and its characteristics. This necessitates further research on the application of different types of sensors and combinations of them in the analysis of food and agricultural products. This obtained results has the prospect of direct application of data from visual images, spectral, ultrasonic and dielectric characteristics, in combination with the selected classifiers and prediction models in the systems for quality assessment of food and agricultural products.
sensor data fusion, eggs, cucumbers, regression, comparative analysis
Cite this paper
Zlatin Zlatev. (2021) Intelligent Methods and Tools for Sensor Data Combination in the Analysis of Food and Agricultural Products. International Journal of Agricultural Science, 6, 188-192