ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry.

Visual Simultaneous Localization and Mapping (V-SLAM) plays a crucial role in the development of intelligent robotics and autonomous navigation systems. However, it still faces significant challenges in handling highly dynamic environments. The prevalent method currently used for dynamic object recognition in the environment is deep learning. However, models such as Yolov5 and Mask R-CNN require significant computational resources, which limits their potential in real-time applications due to hardware and time constraints. To overcome this limitation, this paper proposes ADM-SLAM, a visual SLAM system designed for dynamic environments that builds upon the ORB-SLAM2. This system integrates efficient adaptive feature point homogenization extraction, lightweight deep learning semantic segmentation based on an improved DeepLabv3, and multi-view geometric segmentation. It optimizes keyframe extraction, segments potential dynamic objects using contextual information with the semantic segmentation network, and detects the motion states of dynamic objects using multi-view geometric methods, thereby eliminating dynamic interference points. The results indicate that ADM-SLAM outperforms ORB-SLAM2 in dynamic environments, especially in high-dynamic scenes, where it achieves up to a 97% reduction in Absolute Trajectory Error (ATE). In various highly dynamic test sequences, ADM-SLAM outperforms DS-SLAM and DynaSLAM in terms of real-time performance and accuracy, proving its excellent adaptability.

Mating patterns and pollen dispersal in a Japanese larch (Larix kaempferi) clonal seed orchard: a case study.

Pollination dynamics highly determines the genetic quality of seed orchard crops. However, there is less research about the effect of mating patterns on seed productivity of orchard crops. So far, clonal seed orchards have been producing genetically improved seedlings used for most Japanese larch (Larix kaempferi (Lamb.) Carr.) plantations in China. In the present study, a total of 17 highly variable simple sequence repeat (SSR) markers were used for genotyping a progeny trial population consisting of 647 open-pollinated progenies germinated from seeds which were collected from 63 maternal clones with 140 potential paternal clones in a Japanese larch clonal seed orchard in China. Paternity analysis was used in the present case study in order to evaluate the level of paternal gametic contribution, estimate pollen contamination and selfing rates, and investigate pollination patterns, pollen dispersal patterns and the impact of mating patterns on seed productivity of orchard crops. We observed 93.7% of the success rate of the parental assignment, unequal paternal gametic contribution (0-12.4%) with 6.3% of the progenies derived from pollen contamination or unsampled pollen donors, and absence of evidence for selfing. We also found that pollination rate highly depended on the distance between pollen donors and maternal parents, the majority of the identified crossing (65.7%) occurred between clones within a 150-m radius, and large variations in growth performance existed among the paternal half-siblings. Progeny growth performance (diameter at breast (DBH) and height (HGT)) was measured at Age-20 in order to investigate the impact of mating patterns on timber production of orchard crops. As either the paternal or maternal, two clones (i. e., clones Z38 and Z62) were identified to have produced progenies with higher average stem volume breeding values than that of all of the progenies. Specifically, the genetic gains for volume were 3.53% for the two clones as paternal parents, and 8.26% as the maternal parents at Age-20. Thus, both elite clones were ideal candidates for the construction of next-generation clonal seed orchards due to their synchronous reproductive phenology with greater crossing rate and higher genetic gain. These results improved the pedigree information to provide solid evidence of mating patterns for future design and effective management of seed orchards and for the development of viable long-term breeding strategies for other coniferous species.

Development and characterization of polymorphic genic-SSR markers in Larix kaempferi.

New simple sequence repeat (SSR) markers were developed in the Japanese larch (Larix kaempferi) using unigene sequences for further genetic diversity studies and the genetic improvement of breeding programs. One thousand two handred and thirty five (1235) primer pairs were tested and 165 successfully identified in L. kaempferi. Out of the amplified candidate markers, 145 (90.6%) exhibited polymorphism among 24 individuals of L. kaempferi, with the number of alleles per locus (Na), observed heterozygosity (Ho), expected heterozygosity (He) and polymorphic information content (PIC) averaging at 4.510, 0.487, 0.518 and 0.459, respectively. Cross-species amplification of randomly selection of 30 genic-SSRs among the 145 polymorphic ones showed that 80.0% of the SSRs could be amplified in Larix olgensis, 86.7% could be amplified in Larix principi-rupprechtii and 83.0% could be amplified in Larix gmelinii. High rates of cross-species amplification were observed. The genic-SSRs developed herein would be a valuable resource for genetic analysis of Larix kaempferi and related species, and also have the potential to facilitate the genetic improvement and breeding of larch.