An image stitching method for blades of wind turbine based on background removal preprocessing

Image capturing and stitching system based on scale-invariant feature transform features for untextured and smooth objects

“…In response, the progressive sample consistency (PROSAC) algorithm was proposed to eliminate false matches and enhance the accuracy in matching feature points. Li et al 18 . proposed a stitching technique with background removal preprocessing for images of wind turbine blades.…”

“…In response, the progressive sample consistency (PROSAC) algorithm was proposed to eliminate false matches and enhance the accuracy in matching feature points. Li et al 18 proposed a stitching technique with background removal preprocessing for images of wind turbine blades. A blade has much fewer textural characteristics than the background does in a scenery image with a wind turbine, supporting background-based stitching over stitching based on the blade, leading to error stitching in the image of the blade.…”

A Novel Algorithm for Pre-Processing of WT Blade Images by SS and Bilateral Filter with Machine Learning Frameworks

“…It was considering the position offset of the target to the center of the local window which method designs a CNN with not only a classification head but also a regression head to achieve the precise location of the target for UAV detection. Moreover, based on background removal a preprocessing method is proposed in [8] where in order to overcome the drawbacks of image blurring occur at WT blades. The idea of the method is to carry out image segmentation to get the background part, then the intensities of pixels in background part are replaced by a same value.…”