The use of UAVs to detect litchi in natural environments is imperative for rapid litchi yield estimation and automated harvesting systems. However, UAV-based lychee fruit detection bottlenecks arise from complex canopy architecture and leaf occlusion. This study proposed a collaborative optimization strategy integrating agronomic technique with deep learning. Three leaf thinning intensities (0, 6, and 12 compound leaves) were applied at the early stage of fruit to systematically evaluate their effects on fruit growth, canopy structure, and detection performance. Results indicated that moderate defoliation (six leaves) significantly enhanced canopy openness and light penetration without adversely impacting on yield and fruit quality. Subsequent UAV-based detection under moderate versus no defoliation treatment revealed that the YOLOv8-based model achieved significant performance gains: mean average precision (mAP) increased from 0.818 to 0.884, and the F1-score improved from 0.796 to 0.842. The study contributes a novel collaborative optimization strategy that effectively mitigates occlusion issues in fruit detection. This approach demonstrates that agronomic techniques can be strategically used to enhance AI perception, offering a significant step forward in the integration of agricultural machinery and agronomy for intelligent orchard systems.

The HB (homeobox) transcription factor family plays important roles in plant growth, development, morphogenesis, and stress responses; however, its involvement in longan (Dimocarpus longan Lour.) fruit energy metabolism remains unclear. In this study, 32 DlHB family members were identified in the longan genome, and their physicochemical properties, phylogenetic relationships, gene structures, conserved motifs, conserved domains, tissue-specific expression patterns, promoter cis-acting elements, and collinearity relationships were systematically analyzed. The DlHB family was classified into four subfamilies, HD-ZIP I–IV, with substantial divergence in structural composition, expression patterns, and putative regulatory features. Our previous work showed that 1.5% chitosan (CTS) treatment improved postharvest longan fruit quality through modulation of energy metabolism, and the corresponding CTS-treatment transcriptome was therefore used here to screen energy-metabolism-associated DlHB candidates. DlHB22 was selected as a representative candidate, and its CTS-responsive expression was independently confirmed by qRT-PCR. Exogenous ATP treatment was then used as an independent physiological validation of the relationship between energy metabolism and postharvest storability; ATP-treated fruit showed reduced deterioration together with higher ATP, ADP, and AMP contents and higher activities of H+-ATPase, Ca²+-ATPase, cytochrome c oxidase (CCO), and succinate dehydrogenase (SDH) at 15 d, although adenylate energy charge (AEC) was lower than in the control. DlHB22 localized predominantly to the nucleus. Heterologous overexpression of DlHB22 in tomato accelerated fruit color transition and ripening progression and altered ATP, ADP, and AMP contents, AEC, and the activities of H+-ATPase, Ca²+-ATPase, CCO, and SDH. Transcriptome analysis of DlHB22-overexpressing tomato fruit revealed broad transcriptional changes in pathways associated with central carbon metabolism, energy metabolism, glutathione metabolism, hormone signaling, and MAPK signaling, and qRT-PCR validation of six representative DEGs was consistent with the RNA-seq trends. Because tomato is climacteric whereas longan is non-climacteric, the heterologous tomato results demonstrate the regulatory potential of DlHB22 but do not establish an identical native ripening pathway in longan. Overall, DlHB22 is best regarded as a candidate transcription factor associated with longan fruit energy metabolism, whose native regulatory mechanism requires direct validation in longan.