Mobile mecavision: automatic plant monitoring system as a precision agriculture solution in plant factories

Abstract

Precision agriculture (PA) is an approach in farming systems based on information and technology to enhance agricultural production. PA is applied in plant factories to produce high-quality vegetables in a controlled environment by implementing an automatic crop growth monitoring system. Optimization of the observation process carried out with computer system techniques used in the monitoring process and assessment of the plant condition. Individual observation of cultivated plants requires a monitoring system and numerous equipment. Therefore, a monitoring system equipped with a mobile mechanism is needed to support the monitoring process of crops individually by a single observer. The objective of this study was to design and implement an automatic plant monitoring system based on mobile mechanisms and computer vision using simple equipment in plant factories. The Mobile Mecavision, a monitoring system consists of a mobility system and an image acquisition system controlled by Raspberry Pi. The components are A4988, NEMA 17, and Logitech C270 HD Webcam. The evaluation for system performance carried out using the extreme point shift system and t-test. The aspect of image functionality evaluated using the Structural Similarity Measure test and the result was obtained an image similarity value of 89% to 92% with a 90% level of confidence. The result of the performance test on the mobility aspect obtained that the system shifted numerically but not statistically. The overall performance of the system indicated that the monitoring system could be applied to monitor the crop growth in plant factories.