Field Crop 2025, Vol.8, No.5, 247-257 http://cropscipublisher.com/index.php/fc 252 Figure 2 Analysis of sample quality compliance rate (Adopted from Chen et al., 2025a) Image caption: T1, T2, T3, and T4 indicate that the sowing date was October 5, October 15, October 25, and November 5, respectively; D1, D2, D3, D4, D5, D6, and D7 indicate that the planting density was 135, 180, 225, 270, 315, 360, 405 × 104 plants ha−1, respectively (Adopted from Chen et al., 2025a) 7.2 AI and modeling approaches to predict optimal sowing rates Some people are now using artificial intelligence (AI) to analyze these remote sensing images and field data. AI models, such as convolutional neural networks (CNN) and long short-term memory networks (LSTM), can analyze the growth of crops from sequences of drone photos and predict how much grain they can produce and how many leaves they can grow (Han et al., 2019; Nevavuori et al., 2020). If you do some experiments in the field and combine them with spatial modeling, you can tell you more accurately which plots of land should be planted more densely and which plots should be planted more sparsely. Doing so can achieve the best results for different plots of land and make more money (Istiak et al., 2023). 7.3 Integration with variable-rate seeding technologies Variable Rate Seeding (VRS) technology adjusts seed density based on the actual conditions of each field, such as soil texture, pH, and previous yield data. VRS combines remote sensing, AI analysis, and field data to determine the most appropriate planting method (Sishodia et al., 2020; Šarauskis et al., 2022). This system uses sensors and mapping tools to find out where yields are limited and then automatically controls the seeding rate of the seed drill.
RkJQdWJsaXNoZXIy MjQ4ODYzNA==