Background For the purpose of utilising hybrid vigour to produce possible hybrids with a suitable level of stability,the knowledge of gene activity and combining ability is a crucial prerequisite before choosing desir...Background For the purpose of utilising hybrid vigour to produce possible hybrids with a suitable level of stability,the knowledge of gene activity and combining ability is a crucial prerequisite before choosing desirable parents.The present study was carried out with six parents crossed in full diallel fashion and generated 30 F1 hybrids.These hybrids were evaluated in two replications in Randomized Block Design at Department of Cotton,TNAU for combining ability and gene action.Diallel analysis was carried out according to Griffing’s method-I(parents + F_(1) + reciprocals) and model-I and Hayman’s graphical approach by using INDOSTAT software.Results Analysis of variance for combining ability indicated that mean square values of GCA,SCA and reciprocals were highly significant for all the traits except for the uniformity index.RG763 and K12 showed highly positively significant GCA effects for most of the yield traits while PA838 and K12 for fibre quality traits,so they were found as best general combiners.PAIG379 × K12 and PDB29 × K12 for yield traits,and PDB29 × PA838,RG763 × PA838,and CNA1007 × RG763 cross combinations for fibre quality traits could be recommended for future breeding programms.Conclusion The results of both Griffing’s and Hayman’s approaches showed that non-additive gene action predominates as SCA variance was bigger than GCA variance,so heterosis breeding is thought to be a more fruitful option for enhancing GCA of many traits.展开更多
为精准识别与分类不同花期杭白菊,满足自动化采摘要求,该研究提出一种基于改进YOLOv8s的杭白菊检测模型-YOLOv8s-RDL。首先,该研究将颈部网络(neck)的C2f(faster implementation of CSP bottleneck with 2 convolutions)模块替换为RCS-O...为精准识别与分类不同花期杭白菊,满足自动化采摘要求,该研究提出一种基于改进YOLOv8s的杭白菊检测模型-YOLOv8s-RDL。首先,该研究将颈部网络(neck)的C2f(faster implementation of CSP bottleneck with 2 convolutions)模块替换为RCS-OSA(one-shot aggregation of reparameterized convolution based on channel shuffle)模块,以提升骨干网络(backbone)特征融合效率;其次,将检测头更换为DyHead(dynamic head),并融合DCNv3(deformable convolutional networks v3),借助多头自注意力机制增强目标检测头的表达能力;最后,采用LAMP(layer-adaptive magnitude-based pruning)通道剪枝算法减少参数量,降低模型复杂度。试验结果表明,YOLOv8s-RDL模型在菊米和胎菊的花期分类中平均精度分别达到96.3%和97.7%,相较于YOLOv8s模型,分别提升了3.8和1.5个百分点,同时权重文件大小较YOLOv8s减小了6 MB。该研究引入TIDE(toolkit for identifying detection and segmentation errors)评估指标,结果显示,YOLOv8s-RDL模型分类错误和背景检测错误相较YOLOv8s模型分别降低0.55和1.26。该研究为杭白菊分花期自动化采摘提供了理论依据和技术支撑。展开更多
文摘Background For the purpose of utilising hybrid vigour to produce possible hybrids with a suitable level of stability,the knowledge of gene activity and combining ability is a crucial prerequisite before choosing desirable parents.The present study was carried out with six parents crossed in full diallel fashion and generated 30 F1 hybrids.These hybrids were evaluated in two replications in Randomized Block Design at Department of Cotton,TNAU for combining ability and gene action.Diallel analysis was carried out according to Griffing’s method-I(parents + F_(1) + reciprocals) and model-I and Hayman’s graphical approach by using INDOSTAT software.Results Analysis of variance for combining ability indicated that mean square values of GCA,SCA and reciprocals were highly significant for all the traits except for the uniformity index.RG763 and K12 showed highly positively significant GCA effects for most of the yield traits while PA838 and K12 for fibre quality traits,so they were found as best general combiners.PAIG379 × K12 and PDB29 × K12 for yield traits,and PDB29 × PA838,RG763 × PA838,and CNA1007 × RG763 cross combinations for fibre quality traits could be recommended for future breeding programms.Conclusion The results of both Griffing’s and Hayman’s approaches showed that non-additive gene action predominates as SCA variance was bigger than GCA variance,so heterosis breeding is thought to be a more fruitful option for enhancing GCA of many traits.
文摘为精准识别与分类不同花期杭白菊,满足自动化采摘要求,该研究提出一种基于改进YOLOv8s的杭白菊检测模型-YOLOv8s-RDL。首先,该研究将颈部网络(neck)的C2f(faster implementation of CSP bottleneck with 2 convolutions)模块替换为RCS-OSA(one-shot aggregation of reparameterized convolution based on channel shuffle)模块,以提升骨干网络(backbone)特征融合效率;其次,将检测头更换为DyHead(dynamic head),并融合DCNv3(deformable convolutional networks v3),借助多头自注意力机制增强目标检测头的表达能力;最后,采用LAMP(layer-adaptive magnitude-based pruning)通道剪枝算法减少参数量,降低模型复杂度。试验结果表明,YOLOv8s-RDL模型在菊米和胎菊的花期分类中平均精度分别达到96.3%和97.7%,相较于YOLOv8s模型,分别提升了3.8和1.5个百分点,同时权重文件大小较YOLOv8s减小了6 MB。该研究引入TIDE(toolkit for identifying detection and segmentation errors)评估指标,结果显示,YOLOv8s-RDL模型分类错误和背景检测错误相较YOLOv8s模型分别降低0.55和1.26。该研究为杭白菊分花期自动化采摘提供了理论依据和技术支撑。
