The invention belongs to the technical field of anomaly recognition, and provides a feed
particle processing method and
system with fermented silkworm excrement as a
raw material, and the method specifically comprises the steps: firstly, arranging an
infrared thermal imager in a conditioner in a feed granulator, collecting and storing
infrared images in real time through the
infrared thermal imager in a low-temperature
granulation process, obtaining an
infrared image sequence, and carrying out the low-temperature
granulation process through the
infrared image sequence; the method comprises the following steps of: acquiring a plurality of infrared images, performing regional division on the infrared images to obtain a plurality of
processing regions, performing gelatinization missing risk analysis on a low-temperature
processing process of each micro-
processing region to obtain a gelatinization null value, and performing gelatinization anomaly identification on the feed according to the gelatinization null value. The local gelatinization
hysteresis phenomenon caused by micromechanisms such as
heat transfer path delay, particle
heat capacity difference and
moisture migration limitation is accurately dealt with through gelatinization deletion risk analysis, and feed particles with the pulverization risk are positioned in the feed subjected to low-
temperature treatment, so that the feed pulverization risk of the
raw material prepared from the fermented silkworm excrement is reduced, and the feed pulverization risk of the
raw material prepared from the fermented silkworm excrement is reduced. The
feed quality is improved.