The lettuce fresh weight estimation through group images was researched to increase the accuracy of online feedback of lettuce information in optimal control of greenhouse climate. Based on the group and single lettuce images, the decrease of over error was investigated through cancellation of positive and negative errors based on group estimation. The influence of occlusion on estimation accuracy was quantified. The improvement of estimation accuracy was realized by introducing a novel loss function with the method of deep learning. The results showed that: 1)Compared with the estimation results of lettuce images in single plant images without occlusion, the determination coefficient(R2) of lettuce fresh weight estimation based on group estimation cropping decreased 0.010 8, the normalized root mean squared error(NRMSE) increased 2.69%, and the mean absolute percentage error(MAPE) decreased 2.36%. Although the estimation was slightly lower, the occlusion of the lettuce group can be close to real production. 2)Although there were occlusion issues in group estimation that led to incomplete cropping, the MAPE in group estimation with the cancellation of positive and negative errors was still 3.49% lower than that in single-lettuce estimation. Therefore, group estimation was a better way of providing feedback on lettuce yield. 3)Based on a more optimized MAPE, mean squared percentage error(MSPE) of the loss function can be further decreased to 8.46% which satisfied the requirements on the estimation accuracy of a “soft sensor”. Considering the reality of greenhouse lettuce production, group estimation is a better way to provide online feedback on lettuce yield to the optimal control of the greenhouse climate. Through optimization methods such as deep learning, the estimation error of lettuce yield can be decreased to within 10%.
LiuS Z. Prospects, problems and solutions of hydroponic vegetable industrialization[J]. Agricultural Engineering Technology, 2017, 37(22): 39-45 (in Chinese)
[5]
ChangC L, ChungS C, FuW L, HuangC C. Artificial intelligence approaches to predict growth, harvest day, and quality of lettuce ( Lactuca sativa L) in a IoT-enabled greenhouse system [J]. Biosystems Engineering, 2021, 212: 77-105
LiuJ C, JiC Y. Application status and prospects of crop growth monitoring [J]. Acta Agriculturae Jiangxi, 2007, 19(3): 17-20, 28 (in Chinese)
[8]
RighiniI, VanthoorB, VerheulM J, NaseerM, MaessenH, PerssonT, StanghelliniC. A greenhouse climate-yield model focussing on additional light, heat harvesting and its validation [J]. Biosystems Engineering, 2020, 194: 1-15
[9]
LinD, ZhangL J, XiaX H. Model predictive control of a Venlo-type greenhouse system considering electrical energy, water and carbon dioxide consumption [J]. Applied Energy, 2021, 298: 117163
[10]
XuD, LiY F, DaiA G, ZhaoS M, SongW T. Closed-loop optimal control of greenhouse cultivation based on two-time-scale decomposition: A simulation study in Lhasa [J]. Agronomy, 2022, 13(1): 102
[11]
van BeverenP J M, BontsemaJ, van StratenG, van HentenE J. Minimal heating and cooling in a modern rose greenhouse[J]. Applied Energy, 2015, 137: 97-109
[12]
van HentenE J. Greenhouse climate management: an optimal control approach[D]. Wageningen: Wageningen University and Research, 1994
[13]
XuD, DuS F, van WilligenburgG. Double closed-loop optimal control of greenhouse cultivation[J]. Control Engineering Practice, 2019, 85: 90-99
[14]
BumgarnerN R, MillerW S, KleinhenzM D. Digital image analysis to supplement direct measures of lettuce biomass[J]. HortTechnology, 2012, 22(4): 547-555
[15]
LeeJ W. Machine vision monitoring system of lettuce growth in a state-of-the-art greenhouse [J]. Modern Physics Letters B, 2008, 22(11): 953-958
[16]
YehY H F, LaiT C, LiuT Y, LiuC C, ChungW C, LinT T. An automated growth measurement system for leafy vegetables[J]. Biosystems Engineering, 2014, 117: 43-50
[17]
JungD H, ParkS H, HanX Z, KimH J. Image processing methods for measurement of lettuce fresh weight[J]. Journal of Biosystems Engineering, 2015, 40(1): 89-93
[18]
ChenW T, YehY H F, LiuT Y, LinT T. An automated and continuous plant weight measurement system for plant factory[J]. Frontiers in Plant Science, 2016, 7: 392
[19]
JiangJ S, KimH J, ChoW J. On-the-go image processing system for spatial mapping of lettuce fresh weight in plant factory [J]. IFAC-PapersOnLine, 2018, 51(17): 130-134
[20]
NaganoS, MoriyukiS, WakamoriK, MinenoH, FukudaH. Leaf-movement-based growth prediction model using optical flow analysis and machine learning in plant factory [J]. Frontiers in Plant Science, 2019, 10: 227
[21]
Reyes-YanesA, MartinezP, AhmadR. Real-time growth rate and fresh weight estimation for little gem romaine lettuce in aquaponic grow beds[J]. Computers and Electronics in Agriculture, 2020, 179: 105827
[22]
HuY, WangL, XiangL R, WuQ, JiangH Y. Automatic non-destructive growth measurement of leafy vegetables based on kinect[J]. Sensors, 2018, 18(3): 806
[23]
MortensenA K, BenderA, WhelanB, BarbourM M, SukkariehS, KarstoftH, GislumR. Segmentation of lettuce in coloured 3D point clouds for fresh weight estimation[J]. Computers and Electronics in Agriculture, 2018, 154: 373-381
[24]
ZhangL X, XuZ Y, XuD, MaJ C, ChenY Y, FuZ T. Growth monitoring of greenhouse lettuce based on a convolutional neural network[J]. Horticulture Research, 2020, 7: 124
CuiT Y, YangQ C, ZhangY, XuD, MaJ C. Lettuce fresh weight estimation based on transfer learning and convolutional neural network[J]. Journal of China Agricultural University, 2022, 27(11): 197-206 (in Chinese)
[27]
van HentenE J, BontsemaJ. Non-destructive crop measurements by image processing for crop growth control[J]. Journal of Agricultural Engineering Research, 1995, 61(2): 97-105
[28]
LiX, WangY, FuL. Monitoring lettuce growth using K-means color image segmentation and principal component analysis method[J]. Transactions of the Chinese Society of Agricultural Engineering, 2016, 32(12): 179-186
[29]
HeK M, ZhangX Y, RenS Q, SunJ. Deep residual learning for image recognition[C]. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway, N.J.: IEEE, 2016: 770-778
TanM X, LeQ V. EfficientNet: Rethinking model scaling for convolutional neural networks[C]. In: 36th International Conference on Machine Learning. Long Beach, California: PMLR 97, 2019: 6105-6114
WangW S. Coefficient of variation-a simple and useful statistical index to measure the degree of dispersion[J]. China Statistics, 2007(6): 41-42 (in Chinese)
[34]
RamachandramD, TaylorG W. Deep multimodal learning: A survey on recent advances and trends[J]. IEEE Signal Processing Magazine, 2017, 34(6): 96-108
[35]
BuxbaumN, LiethJ H, EarlesM. Non-destructive plant biomass monitoring with high spatio-temporal resolution via proximal RGB-D imagery and end-to-end deep learning [J]. Frontiers in Plant Science, 2022, 13: 758818
[36]
LinZ X, FuR M, RenG Q, ZhongR H, YingY B, LinT. Automatic monitoring of lettuce fresh weight by multi-modal fusion based deep learning[J]. Frontiers in Plant Science, 2022, 13: 980581
[37]
ZhangQ J, ZhangX Y, WuY L, LiX S. TMSCNet: A three-stage multi-branch self-correcting trait estimation network for RGB and depth images of lettuce [J]. Frontiers in Plant Science, 2022, 13: 982562