UAVLitchi: A UAV-based benchmark dataset for litchi fruit segmentation and detection in natural environment

Key Points

• Litchi fruit segmentation and detection accuracy improved significantly through the use of the UAV dataset.
• The dataset consists of over 4,000 annotated images captured in various natural environments to ensure comprehensive coverage.
• UAV image data collection method was utilized to create a benchmark for segmentation tasks in agriculture and computer vision.
• This study highlights the need for standardized datasets for enhanced algorithm performance in real-world applications.