The invention discloses a potassium feldspar ceramic sintering whiteness detection system and method, and relates to the technical field of ceramic detection.The method comprises the steps that a space coordinate system is constructed in a ceramic sintering furnace, a plurality of high-temperature industrial cameras are used for periodically capturing ceramic images, image pixel coordinates are converted into three-dimensional coordinates through preprocessing, and then ceramic pixel points are screened out; with reference to historical sintering data, drawing a time sequence chart of the ceramic surface whiteness changing along with time, determining a standard whiteness value of each monitoring period, and meanwhile, fitting a correlation function of the in-furnace temperature and the sintering time; extracting RGB color data from the ceramic image data of the current monitoring period, converting the RGB color data into a whiteness value, and establishing a real-time monitoring whiteness time sequence diagram; by comparing the actual whiteness with the standard whiteness, the ceramic pixel points are classified, the abnormal pixels are identified, and whether the temperature needs to be adjusted or whether the model is abnormal or not is intelligently judged according to the number and the type of the abnormal pixels, so that the stability and the consistency of the ceramic sintering quality are ensured.