Background The lint percentage of seed cotton is one of the most important parameters for evaluating seed cotton quality and affects its price.The traditional measuring method of lint percentage is labor-intensive and...Background The lint percentage of seed cotton is one of the most important parameters for evaluating seed cotton quality and affects its price.The traditional measuring method of lint percentage is labor-intensive and time-consuming;thus,an efficient and accurate measurement method is needed.In recent years,classification-based deep learning and computer vision have shown promise in solving various classification tasks.Results In this study,we propose a new approach for detecting the lint percentage using MobileNetV2 and transfer learning.The model is deployed on a lint percentage detection instrument,which can rapidly and accurately determine the lint percentage of seed cotton.We evaluated the performance of the proposed approach using a dataset comprising 66924 seed cotton images from different regions of China.The results of the experiments showed that the model with transfer learning achieved an average classification accuracy of 98.43%,with an average precision of 94.97%,an average recall of 95.26%,and an average F1-score of 95.20%.Furthermore,the proposed classification model achieved an average accuracy of 97.22%in calculating the lint percentage,showing no significant difference from the performance of experts(independent-sample t-test,t=0.019,P=0.860).Conclusion This study demonstrated the effectiveness of the MobileNetV2 model and transfer learning in calculating the lint percentage of seed cotton.The proposed approach is a promising alternative to traditional methods,providing a rapid and accurate solution for the industry.展开更多
To introgress the good fiber quality and yield from Gossypium barbadense into a commercial Upland cotton variety, a high‐density simple sequence repeat (SSR) genetic linkage map was developed from a BC1F1 populatio...To introgress the good fiber quality and yield from Gossypium barbadense into a commercial Upland cotton variety, a high‐density simple sequence repeat (SSR) genetic linkage map was developed from a BC1F1 population of Gossypium hirsutum × Gossypium barbadense. The map com-prised 2,292 loci and covered 5115.16 centiMorgan (cM) of the cotton AD genome, with an average marker interval of 2.23 cM. Of the marker order for 1,577 common loci on this new map, 90.36% agrees well with the marker order on the D genome sequence genetic map. Compared with five pub-lished high‐density SSR genetic maps, 53.14% of marker loci were newly discovered in this map. Twenty‐six quantitative trait loci (QTLs) for lint percentage (LP) were identified on nine chromosomes. Nine stable or common QTLs could be used for marker‐assisted selection. Fifty percent of the QTLs were from G. barbadense and increased LP by 1.07%–2.41%. These results indicated that the map could be used for screening chromosome substitution segments from G. barbadense in the Upland cotton background, identifying QTLs or genes from G. barbadense, and further developing the gene pyramiding effect for improving fiber yield and quality.展开更多
基金National Natural Science Foundation of China(Grant number:11904327,61905223,and 62073299)Training Plan of Young Backbone Teachers in Universities of Henan Province(2023GGJS087)+1 种基金Henan Provincial Science and Technology Research Project(222102110279,222102210085,and 242102210157)Project of Central Plains Science and Technology Innovation Leading Talents(224200510026).
文摘Background The lint percentage of seed cotton is one of the most important parameters for evaluating seed cotton quality and affects its price.The traditional measuring method of lint percentage is labor-intensive and time-consuming;thus,an efficient and accurate measurement method is needed.In recent years,classification-based deep learning and computer vision have shown promise in solving various classification tasks.Results In this study,we propose a new approach for detecting the lint percentage using MobileNetV2 and transfer learning.The model is deployed on a lint percentage detection instrument,which can rapidly and accurately determine the lint percentage of seed cotton.We evaluated the performance of the proposed approach using a dataset comprising 66924 seed cotton images from different regions of China.The results of the experiments showed that the model with transfer learning achieved an average classification accuracy of 98.43%,with an average precision of 94.97%,an average recall of 95.26%,and an average F1-score of 95.20%.Furthermore,the proposed classification model achieved an average accuracy of 97.22%in calculating the lint percentage,showing no significant difference from the performance of experts(independent-sample t-test,t=0.019,P=0.860).Conclusion This study demonstrated the effectiveness of the MobileNetV2 model and transfer learning in calculating the lint percentage of seed cotton.The proposed approach is a promising alternative to traditional methods,providing a rapid and accurate solution for the industry.
基金funded by the National Basic Research Program of China (973 Project) (2010CB126000)the National High Technology Research and Development Program of China (2012AA101108)+1 种基金the National Natural Science Foundation of China (31101188)the fund project of Director (SJA1203)
文摘To introgress the good fiber quality and yield from Gossypium barbadense into a commercial Upland cotton variety, a high‐density simple sequence repeat (SSR) genetic linkage map was developed from a BC1F1 population of Gossypium hirsutum × Gossypium barbadense. The map com-prised 2,292 loci and covered 5115.16 centiMorgan (cM) of the cotton AD genome, with an average marker interval of 2.23 cM. Of the marker order for 1,577 common loci on this new map, 90.36% agrees well with the marker order on the D genome sequence genetic map. Compared with five pub-lished high‐density SSR genetic maps, 53.14% of marker loci were newly discovered in this map. Twenty‐six quantitative trait loci (QTLs) for lint percentage (LP) were identified on nine chromosomes. Nine stable or common QTLs could be used for marker‐assisted selection. Fifty percent of the QTLs were from G. barbadense and increased LP by 1.07%–2.41%. These results indicated that the map could be used for screening chromosome substitution segments from G. barbadense in the Upland cotton background, identifying QTLs or genes from G. barbadense, and further developing the gene pyramiding effect for improving fiber yield and quality.