OPAL-Flow: Orientation-aware rice panicle detection and minute-scale anthesis rhythm identification under field conditions

Accurate timing of rice panicle anthesis is critical for quantifying sterility risk under heat and humidity, yet minute-scale field measurement remains challenging because anthesis is transient and spikelets are tiny and difficult to detect. To address this, we present OPAL-Flow, a pipeline that provides single-panicle anthesis start and peak times under field conditions, consisting of a detector for panicle detection and tracking, a super-resolution reconstruction model, and an event-time pinpointing model. For slender panicle detection and panicle pose normalization, YOLO-SnakePanNet was introduced by using Dynamic Snake Convolution with a lightweight box-rotated head. Ablation experiments show that YOLO-SnakePanNet achieved mAP@50 of 94.4%, improving by 3.6% over the YOLOv11 while reducing computation by 0.7 GFLOPs. For panicle-level anthesis pinpointing, PanicleTimeMAE was proposed by incorporating a pyramid-dilated temporal convolutional network and a confidence-aware smoothing gate into the transformer, reaching Acc@±1 of 0.85 on 5-min sampled sequences (±1 frame = ±5 min), yielding a 40% decrease in MAE over VideoMAEv2. Finally, correlation analysis between variety-level anthesis start time (T_start) and peak time (T_peak) and same-day meteorology showed that higher photosynthetically active radiation (r = −0.543/−0.573 for T_start/T_peak) and temperature (r = −0.288/−0.272) advanced anthesis, whereas higher relative humidity (r = 0.397/0.438) and rainfall (r = 0.428/0.502) delayed anthesis. The variance decomposition within fixed-effects model for (= 0.651) and (= 0.648) prediction shows that variance mainly attributed to meteorological effects (64%) and variety effects (33.5%). Overall, OPAL-Flow enables variety selection for heat- and humidity-resilient anthesis in rice breeding and supports ecophysiological dissection of anthesis regulation.