The application discloses a kind of based on image recognition's demolition garbage aggregate classification method, belong to demolition garbage
processing technical field.The application collects the visible light image of aggregate,
ultraviolet induced
fluorescence image and
infrared thermal image under pulse thermal excitation, shape class and shape irregularity index are extracted from visible light image,
mortar coverage is extracted from
ultraviolet fluorescence image, and microcrack density is extracted from
infrared thermal image;Then determine shape grade and defect grade, fuse and output first to fourth level classification
label.The application also identifies red
brick particles through near-
infrared reflectance spectroscopy, and distinguishes demolition induced cracks from original microcracks using
thermal decay characteristics.The application realizes comprehensive, rapid and automated evaluation of aggregate shape, surface
mortar, internal cracks and impurities, and the classification results can directly guide the utilization of aggregate resources, effectively improving the
utilization rate of demolition garbage aggregate with high added value.