A survey of few-shot learning in smart agriculture: developments, applications, and challenges

Yang, Jiachen; Guo, Xiaolan; Li, Yang; Marinello, Francesco; Erċışlı, Sezai; Zhang, Zhuo

doi:10.1186/s13007-022-00866-2

Cited by 100 publications

(57 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When the target point or obstacles change slightly, these methods need to re-plan because of the lack of adaptability and flexibility. With the rapid development of deep learning in recent years, intelligent methods have played an increasingly important role in smart agriculture [15,16], image screening [17], environmental monitoring [18], edge computing [19], and path planning [20]. At the same time, it provides an effective technical means for agricultural production, automatic driving, and other fields.…”

Section: Introductionmentioning

confidence: 99%

The Intelligent Path Planning System of Agricultural Robot via Reinforcement Learning

Yang

Chen

2022

Sensors

Self Cite

View full text Add to dashboard Cite

Agricultural robots are one of the important means to promote agricultural modernization and improve agricultural efficiency. With the development of artificial intelligence technology and the maturity of Internet of Things (IoT) technology, people put forward higher requirements for the intelligence of robots. Agricultural robots must have intelligent control functions in agricultural scenarios and be able to autonomously decide paths to complete agricultural tasks. In response to this requirement, this paper proposes a Residual-like Soft Actor Critic (R-SAC) algorithm for agricultural scenarios to realize safe obstacle avoidance and intelligent path planning of robots. In addition, in order to alleviate the time-consuming problem of exploration process of reinforcement learning, this paper proposes an offline expert experience pre-training method, which improves the training efficiency of reinforcement learning. Moreover, this paper optimizes the reward mechanism of the algorithm by using multi-step TD-error, which solves the probable dilemma during training. Experiments verify that our proposed method has stable performance in both static and dynamic obstacle environments, and is superior to other reinforcement learning algorithms. It is a stable and efficient path planning method and has visible application potential in agricultural robots.

show abstract

Section: Introductionmentioning

confidence: 99%

The Intelligent Path Planning System of Agricultural Robot via Reinforcement Learning

Yang

Chen

2022

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…Based on image information, the parameters including crop counting, stem diameter, and stem height can be measured. However, there are still have some problems, such as difficulty in image segmentation, influence of illumination, lack of 3D information, requiring massive data calibration [ 23 – 25 ], improving the accuracy of measurement, and so on.…”

Section: Introductionmentioning

confidence: 99%

Banana plant counting and morphological parameters measurement based on terrestrial laser scanning

et al. 2022

View full text Add to dashboard Cite

Background The number of banana plants is closely related to banana yield. The diameter and height of the pseudo-stem are important morphological parameters of banana plants, which can reflect the growth status and vitality. To address the problems of high labor intensity and subjectivity in traditional measurement methods, a fast measurement method for banana plant count, pseudo-stem diameter, and height based on terrestrial laser scanning (TLS) was proposed. Results First, during the nutritional growth period of banana, three-dimensional (3D) point cloud data of two measured fields were obtained by TLS. Second, the point cloud data was preprocessed. And the single plant segmentation of the canopy closed banana plant point cloud was realized furtherly. Finally, the number of banana plants was obtained by counting the number of pseudo-stems, and the diameter of pseudo-stems was measured using a cylindrical segmentation algorithm. A sliding window recognition method was proposed to determine the junction position between leaves and pseudo-stems, and the height of the pseudo-stems was measured. Compared with the measured value of artificial point cloud, when counting the number of banana plants, the precision,recall and percentage error of field 1 were 93.51%, 94.02%, and 0.54% respectively; the precision,recall and percentage error of field 2 were 96.34%, 92.00%, and 4.5% respectively; In the measurement of pseudo-stem diameter and height of banana, the root mean square error (RMSE) of pseudo-stem diameter and height of banana plant in field 1 were 0.38 cm and 0.2014 m respectively, and the mean absolute percentage error (MAPE) were 1.30% and 5.11% respectively; the RMSE of pseudo-stem diameter and height of banana plant in field 2 were 0.39 cm and 0.2788 m respectively, and the MAPE were 1.04% and 9.40% respectively. Conclusion The results show that the method proposed in this paper is suitable for the field measurement of banana count, pseudo-stem diameter, and height and can provide a fast field measurement method for banana plantation management.

show abstract

“…Studies should focus on the laws of crop growth and development, reveal gene regulation pathways and optimise precision management of crop cultivation and acceleration of crop improvement [ 5 ]. The phenotype acquisition platform is an important hardware basis for rapid screening of germplasm resources, phenotype identification, and formation mechanism research [ 6 – 8 ], and it is mainly composed of mechanical devices or drones equipped with sensors [ 9 , 10 ].…”

Section: Introductionmentioning

confidence: 99%

Evaluation of an intelligent artificial climate chamber for high-throughput crop phenotyping in wheat

et al. 2022

View full text Add to dashboard Cite

Background The superposition of COVID-19 and climate change has brought great challenges to global food security. As a major economic crop in the world, studying its phenotype to cultivate high-quality wheat varieties is an important way to increase grain yield. However, most of the existing phenotyping platforms have the disadvantages of high construction and maintenance costs, immobile and limited in use by climatic factors, while the traditional climate chambers lack phenotypic data acquisition, which makes crop phenotyping research and development difficult. Crop breeding progress is slow. At present, there is an urgent need to develop a low-cost, easy-to-promote, climate- and site-independent facility that combines the functions of crop cultivation and phenotype acquisition. We propose a movable cabin-type intelligent artificial climate chamber, and build an environmental control system, a crop phenotype monitoring system, and a crop phenotype acquisition system. Result We selected two wheat varieties with different early vigor to carry out the cultivation experiments and phenotype acquisition of wheat under different nitrogen fertilizer application rates in an intelligent artificial climate chamber. With the help of the crop phenotype acquisition system, images of wheat at the trefoil stage, pre-tillering stage, late tillering stage and jointing stage were collected, and then the phenotypic information including wheat leaf area, plant height, and canopy temperature were extracted by the crop type acquisition system. We compared systematic and manual measurements of crop phenotypes for wheat phenotypes. The results of the analysis showed that the systematic measurements of leaf area, plant height and canopy temperature of wheat in four growth periods were highly correlated with the artificial measurements. The correlation coefficient (r) is positive, and the determination coefficient (R2) is greater than 0.7156. The root mean square error (RSME) is less than 2.42. Among them, the crop phenotype-based collection system has the smallest measurement error for the phenotypic characteristics of wheat trefoil stage. The canopy temperature RSME is only 0.261. The systematic measurement values of wheat phenotypic characteristics were significantly positively correlated with the artificial measurement values, the fitting degree was good, and the errors were all within the acceptable range. The experiment showed that the phenotypic data obtained with the intelligent artificial climate chamber has high accuracy. We verified the feasibility of wheat cultivation and phenotype acquisition based on intelligent artificial climate chamber. Conclusion It is feasible to study wheat cultivation and canopy phenotype with the help of intelligent artificial climate chamber. Based on a variety of environmental monitoring sensors and environmental regulation equipment, the growth environment factors of crops can be adjusted. Based on high-precision mechanical transmission and multi-dimensional imaging sensors, crop images can be collected to extract crop phenotype information. Its use is not limited by environmental and climatic factors. Therefore, the intelligent artificial climate chamber is expected to be a powerful tool for breeders to develop excellent germplasm varieties.

show abstract

A survey of few-shot learning in smart agriculture: developments, applications, and challenges

Cited by 100 publications

References 53 publications

The Intelligent Path Planning System of Agricultural Robot via Reinforcement Learning

The Intelligent Path Planning System of Agricultural Robot via Reinforcement Learning

Banana plant counting and morphological parameters measurement based on terrestrial laser scanning

Evaluation of an intelligent artificial climate chamber for high-throughput crop phenotyping in wheat

Contact Info

Product

Resources

About