A lightweight deep learning approach for dual-side walnut kernel quality inspection via heterogeneous feature fusion

Key Points

• This research aims to explore a lightweight deep learning method for inspecting walnut kernel quality.
• Developed a deep learning model to inspect walnut kernels.
• Utilized feature fusion techniques for quality assessment.
• Targeted dual-side evaluation of walnut kernels.
• Achieved enhanced quality inspection accuracy for walnut kernels.
• Demonstrated efficiency in processing through a lightweight model.