AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden.

The assessment of food and industrial crops during harvesting is important to determine the quality and downstream processing requirements, which in turn affect their market value. While machine learning models have been developed for this purpose, their deployment is hindered by the high cost of labelling the crop images to provide data for model training. This study examines the capabilities of semi-supervised and active learning to minimise effort when labelling cotton lint samples while maintaining high classification accuracy. Random forest classification models were developed using supervised learning, semi-supervised learning, and active learning to determine Egyptian cotton grade. Compared to supervised learning (80.20-82.66%) and semi-supervised learning (81.39-85.26%), active learning models were able to achieve higher accuracy (82.85-85.33%) with up to 46.4% reduction in the volume of labelled data required. The primary obstacle when using machine learning for Egyptian cotton grading is the time required for labelling cotton lint samples. However, by applying active learning, this study successfully decreased the time needed from 422.5 to 177.5 min. The findings of this study demonstrate that active learning is a promising approach for developing accurate and efficient machine learning models for grading food and industrial crops.

Impact of Plant Spacing and Nitrogen Rates on Growth Characteristics and Yield Attributes of Egyptian Cotton (Gossypium barbadense L.).

This current study was performed to determine the influences of plant spacing, Nitrogen (N) fertilization rate and their effect, on growth traits, yield, and yield components of cotton (Gossypium barbadense L.) cv. Giza 97 during the 2019 and 2020 seasons. A split plot experiment in three replicates was utilized whereas the cotton seeds were planted at 20, 30, and 40 cm, as main plots and nitrogen at 75, 100, and 125%, was in subplots. The results revealed that the planting spacing at 40 cm significantly (p ≤ 0.01) increased plant height, number of fruiting branches per plant, number of bolls per plant, boll weight (BW), lint percentage (L%), seed cotton yield (SCY), lint cotton yield (LCY), seed index and lint index by 165.68 cm, 20.92, 23.93, 3.75 g, 42.01%, 4.24 ton/ha, 5.16 ton/ha, 12.05, 7.86, respectively, as average in both seasons. The application of N fertilizer rate at 125% caused a maximum increase in growth and yield parameters i.e., plant height (169.08 cm), number of vegetative branches (2.67), number of fruiting branches per plant (20.82), number bolls per fruiting branch (1.39), number of bolls per plant (23.73), boll weight (4.1 g), lint percent (41.9%), seed index (11.8 g), and lint index (8.2), while the plants treated with 100% N rates exhibited highest seed cotton yield (4.3 ton/ha) and lint cotton yield (5.6 ton/ha), as average in both seasons. Combining plant spacing at 40 cm between plants with a 100% N fertilizer rate recorded the highest lint cotton yield (5.67 ton/ha), while the highest seed cotton yield (4.43 and 4.50 ton/ha) was obtained from 125% N fertilizer rate under planting spacing 20 and 40 cm, respectively. Conclusively, a wide density (40 cm) with 125% N is a promising option for improved biomass, cotton growth, yield, physiological traits, and fiber quality.

Zinc Nutrition Responses to Agronomic and Yield Traits, Kernel Quality, and Pollen Viability in Rice (Oryza sativa L.).

Rice (Oryza sativa L.) is one of the major cereal crops worldwide with wheat and maize. A total of two field experiments were performed to evaluate the response of some rice cultivars to various foliar zinc (Zn) concentrations based on different measurements, such as agronomic, yield, yield compounds, and grain technological parameters. The experimental layout was a split plot in three replicates; the five rice cultivars (Skaha 101, Giza178, Yasmeen, Fourate, and Amber 33) were distributed in the main plots while the four foliar applications of Zn (1,500, 2,000, 2,500 mg/L besides spray water) were occupied the sub-plots. The findings showed significant differences among the five rice cultivars regarding plant height, grain yield, straw yield, biological yield, harvest index, 1,000-grain weight, panicle length, protein percentage, and grain Zn content. There is a significant effect of Zn on all plant attributes. A significant interaction between rice cultivars and foliar application of Zn was observed, whereas fertilizing Giza 178 with foliar application of Zn at the rate of 2,500 mg/L achieved the highest mean values of grain yield and straw yield, biological yield, harvest index, 1,000-grain weight, panicle length, protein %, and Zn content followed by Sakha 101 with Zn application at the rate of 2,000 mg/L, respectively, in both seasons. The rice cultivars significantly differed in hulling (%), broken (%), hardness, grain length, shape, amylose (%), gel consistency, and gelatinization temperature. Unfortunately, the commercial Zn product used was genotoxic to pollen grains with a higher rate of Zn. Aberrations were observed such as stickiness, ultrastructural changes in the exterior and interior walls, partially or fully degenerated grains, and shrunken and unfilled grains. This study concluded that using Zn application at the rate of 2,000 mg/L to protect human and environmental health, the side effects and toxicity of the local commercial Zn product market should be investigated before making recommendations to farmers.