Solar energy cell appearance slurry leakage defect detection method

The invention provides a solar energy cell appearance slurry leakage defect detection method. According to slurry leakage defect kind characteristics, classification detection and program algorithm partitioning detection are carried out to realize vision detection on the solar energy cell surface appearance slurry leakage defect, the defect detection method comprises 11 parts, including gray conversion, rotation transformation, ROI segmentation, conversion to be HSI space images, laplace's transformation, mean value filtering, image binarization, morphological transformation, convolutional processing, connected domain selection and defect characteristic area screening, the surface slurry leakage defect detection method further comprises four main parts, the first part is for image pre-processing, and an original image is converted into the processing-convenient image information; the second part is for removing main and auxiliary grid lines, and slurry leakage white points outside thegrid lines are detected;