Accurate plant counting supports stand assessment and crop management, but canopy overlap, camera motion, and temporary occlusion complicate video-based monitoring. This study developed an RGB image- and video-based framework for detecting, tracking, and counting cabbage seedlings and grown plants under greenhouse and open-field conditions. The framework integrated YOLOv8n with convolutional block attention modules (CBAM) and an improved BoT-SORT procedure incorporating trajectory persistence, short-gap relinking, class-consistent identity management, and counting-event filtering. Detector performance was evaluated across five independent training runs, counting performance across nine held-out videos, and tracking performance on four manually annotated videos. YOLOv8n-CBAM achieved mean precision, recall, mAP@0.50, and mAP@0.50:0.95 values of 0.880 ± 0.007, 0.920 ± 0.007, 0.960 ± 0.006, and 0.600 ± 0.008, respectively (mean ± sample standard deviation). The complete framework achieved 99.27% mean counting accuracy, a mean absolute error of 0.56 plants, a root mean square error of 0.88 plants, and an R² of 0.997. On the annotated subset, Improved BoT-SORT increased mean IDF1 from 79.6% to 82.3% and mean HOTA from 74.3% to 83.2%, while reducing total identity switches from 23 to 12. Complete-pipeline throughput was 23.38 frames per second on an RTX 3090 workstation. These findings support cabbage population monitoring under the evaluated conditions; independent-site validation, tracking-component ablation, and deployment-device testing remain necessary to establish broader applicability.

Reliable actuator operation is essential for maintaining stable environmental and production conditions in smart vertical farms. The objective of this study was to characterize power consumption data from five actuator types, including a brushless direct-current (BLDC) fan, a dehumidifier, a light-emitting diode (LED) lighting system, a pump, and a solenoid valve, under normal and abnormal operating conditions using a common experimental approach. The combined evaluation of these five actuator types provided a unified basis for comparing power-consumption characteristics across actuators with different operating mechanisms. The experiments were conducted under controlled laboratory-scale conditions, and each operating condition was evaluated in three experimental repetitions. Power consumption data were measured using Hall-effect current sensors at a sampling frequency of 1 kHz. Time-domain characteristics, including power magnitude, variability, power range, coefficient of variation (CV), transient behavior, and temporal variability, were evaluated together with standardized feature heatmaps and relative-frequency distributions. The results indicated that power responses varied according to actuator type and abnormal condition. Blockage or restricted operation increased power demand and variability in the BLDC fan, pump, and solenoid valve, whereas pump dry-run operation produced a lower operating demand. Dehumidifier short cycling resulted in repeated activation and interruption and produced pronounced signal variability, with a CV of 31.3% and a power range of 598.2 W. Unstable LED flicker altered both power level and temporal behavior, while intermittent connections produced repeated power reductions in the pump and solenoid valve. No-response conditions remained near actuator-specific baseline levels, with mean power ranging from approximately 0.5 to 6.0 W. These results confirmed that actuator operational status could be represented by combinations of power magnitude, variability, transient response, and temporal behavior. The proposed power-consumption characterization provides a basis for actuator-status monitoring and future fault-detection applications in smart vertical farms.