Selective transplantation method of leafy vegetable seedlings based on ResNet 18 network

Area of Science:

• Agricultural technology
• Computer vision
• Machine learning

Background:

• Low survival rates in transplanted leafy vegetables are a significant agricultural challenge.
• Current seedling selection methods are often manual and inefficient.

Purpose of the Study:

• To develop an automated selective transplanting method for leafy vegetable seedlings.
• To improve the survival rate of transplanted seedlings using artificial intelligence.

Main Methods:

• A dataset of 3,388 lettuce seedling images was created.
• The ResNet 18 network was employed for transfer learning, classification, and screening.
• Background removal technology was used to optimize image processing time.

Main Results:

• The ResNet 18 model achieved a screening accuracy of 97.44% on the validation set.
• Precision and recall rates for healthy and unhealthy seedlings were high (e.g., 97.56% precision for healthy seedlings).
• The model processed single images in just 0.0129 seconds.

Conclusions:

• The proposed ResNet 18-based selective transplanting method significantly enhances seedling screening accuracy.
• This AI-driven approach offers a more efficient and effective alternative to traditional methods.
• The technology provides valuable support for improving leafy vegetable cultivation and survival rates.