Lettuce Production in Intelligent Greenhouses—3D Imaging and Computer Vision for Plant Spacing Decisions

Abstract

Recent studies indicate that food demand will increase by 35–56% over the period 2010–2050 due to population increase, economic development, and urbanization. Greenhouse systems allow for the sustainable intensification of food production with demonstrated high crop production per cultivation area. Breakthroughs in resource-efficient fresh food production merging horticultural and AI expertise take place with the international competition “Autonomous Greenhouse Challenge”. This paper describes and analyzes the results of the third edition of this competition. The competition’s goal is the realization of the highest net profit in fully autonomous lettuce production. Two cultivation cycles were conducted in six high-tech greenhouse compartments with operational greenhouse decision-making realized at a distance and individually by algorithms of international participating teams. Algorithms were developed based on time series sensor data of the greenhouse climate and crop images. High crop yield and quality, short growing cycles, and low use of resources such as energy for heating, electricity for artificial light, and CO2 were decisive in realizing the competition’s goal. The results highlight the importance of plant spacing and the moment of harvest decisions in promoting high crop growth rates while optimizing greenhouse occupation and resource use. In this paper, images taken with depth cameras (RealSense) for each greenhouse were used by computer vision algorithms (Deepabv3+ implemented in detectron2 v0.6) in deciding optimum plant spacing and the moment of harvest. The resulting plant height and coverage could be accurately estimated with an R2 of 0.976, and a mIoU of 98.2, respectively. These two traits were used to develop a light loss and harvest indicator to support remote decision-making. The light loss indicator could be used as a decision tool for timely spacing. Several traits were combined for the harvest indicator, ultimately resulting in a fresh weight estimation with a mean absolute error of 22 g. The proposed non-invasively estimated indicators presented in this article are promising traits to be used towards full autonomation of a dynamic commercial lettuce growing environment. Computer vision algorithms act as a catalyst in remote and non-invasive sensing of crop parameters, decisive for automated, objective, standardized, and data-driven decision making. However, spectral indexes describing lettuces growth and larger datasets than the currently accessible are crucial to address existing shortcomings between academic and industrial production systems that have been encountered in this work.

Monitoring of a Productive Blue-Green Roof Using Low-Cost Sensors

“…Here, in direct sunlight, both 99th and 98th percentile RMSE_B values are less than 7 cm, which appears to be better than the findings of Neupane et al [ 44 ]. Petropoulou et al deployed a RealSense camera in their study for estimating lettuce growth with an R 2 of 0.976 [ 18 ]. In this study, the okra FPH and IPH were found to be significantly correlated, with R 2 values for the 98th and 99th percentiles of 0.9729 ( Figure 7 ) and 0.9711, respectively.…”

“…IoT facilitates remote monitoring and operation, requires minimal energy, automatically records large volumes of data, may be controlled remotely, and improves efficiency [ 15 , 16 , 17 ]. IoT-based farming can promote agricultural automation by equipping it with a sensor network [ 18 ], which could increase crop yield and reduce operating costs [ 19 ]. During the crop growth period, observations and analyses can be conducted even in poor weather conditions when routine checks are not feasible.…”

“…However, 2D imaging has several limitations such as inconsistency in the physical dimension and distance from the image sensor and the inability to fully extract 3D information due to complicated plant structures and overlying foliage when evaluated against 3D imaging [ 31 ]. Crop growth rate employing image segmentation algorithms renders potential, although experiments are often conducted in semi-commercial settings, emphasizing the importance of high-quality image settings [ 18 ]. It is essential to develop low-cost technology for image-based crop phenotyping on open rooftops.…”

Smart Glazing for Energy- and Cost-Efficient Greenhouse Humidity Regulation

“…Implementing a more energy-efficient approach to humidity regulation can significantly reduce the energy consumption associated with agriculture, which currently amounts to as much as 1872 trillion Btu per year in the United States . In recent years, new technologies including functional greenhouse covers and Internet of Things have been developed to enable energy-independent, labor-independent, and high-yield greenhouses. − Major efforts have been devoted to intelligent photovoltaic systems that regulate solar radiation, generate energy, and control temperature. , However, advanced solutions for humidity regulation have not been explored in detail.…”

Development of a machine vision-based weight prediction system of butterhead lettuce (Lactuca sativa L.) using deep learning models for industrial plant factory

“…Notably, plant factories must track the fresh weight of crops because they are cultured in bulk in the system. This is because the owner should meet the ordered quantity and quality, including certain sizes and weights, and have a specific date of shipping when an indoor farm is commercialized ( Petropoulou et al., 2023 ). A delay in shipment is one of the most severe problems in commercial indoor farming.…”