LE-YOLO: Lightweight and Efficient Detection Model for Wind Turbine Blade Defects Based on Improved YOLO

Aim:

        To develop a lightweight and efficient detection model using YOLO-v8 for identifying wind turbine blade defects with improved accuracy and real-time performance.

Abstract:

          A novel approach for detecting small targets of wind turbine blade surface damage, particularly minute defects within intricate scenes depicted in low-resolution images, is introduced through the utilization of an enhanced YOLO-v8 algorithm. The method incorporates advanced techniques to improve detection accuracy without compromising inference speed, tailored to the specific characteristics of wind turbine blade defects. Experimental findings demonstrate that the refined algorithm surpasses the existing YOLO-v7 approach in terms of accuracy and efficiency, facilitating more effective detection of wind turbine blade defects while meeting real-time target detection requisites.

Existing System:

           The existing system utilizes an enhanced YOLO-v7 algorithm to detect small targets of wind turbine blade defects. Key improvements in the system include the integration of Ghost Shading Mixing Convolution (GSConv) for accuracy and Simple Attention Mechanism (SimAM) for better detection of small targets. The Edge Intersection over Union (EIoU) is employed as the edge loss function to expedite convergence. The system achieves an average accuracy of 78.7% .

Problem Definition:

         Despite the advancements in YOLO-v7, the system faces challenges in detecting intricate and minute defects on wind turbine blades, particularly in low-resolution images. There is a need for a more accurate and efficient model to improve defect detection while maintaining real-time processing capabilities.

Proposed System:

          The proposed system leverages the YOLO-v8 algorithm to address the shortcomings of the existing system. By incorporating advanced techniques and optimizations, the system aims to achieve improved accuracy and detection capabilities for minute defects on wind turbine blades. YOLO-v8 will integrate enhancements to ensure higher detection accuracy, better handling of intricate scenes, and faster processing speeds suitable for real-time applications.

Advantage:

1. Improved detection accuracy for small and intricate defects.
2. Enhanced real-time performance suitable for practical applications.
3. Lightweight and efficient model design for ease of deployment.

Algorithm: YOLO-v8 with custom optimizations for wind turbine blade defect detection.