Vegetable pests and
diseases are some of the main factors affecting
vegetable yield. Accurate monitoring and intelligent identification
of vegetable pests and diseases are prerequisites for pest forecasting
and integrated control. In this study, a vegetable pest identification
system based on an improved Alexnet algorithm and 5G communication
is designed. The system uses high-definition cameras and 5G communication
modules to form the pest monitoring network. It builds an image recognition
model based on the improved Alexnet algorithm to identify vegetable
pests, and then it collects pictures for transmission to the terminal.
After the experimental test, the pest identification system proposed
in this study accounts for only 11.71, 11.91, 30.92, and 31.38% of
the identification system of the 4G communication network in terms
of transmission delay, transmission jitter, packet loss rate, and
packet error rate, respectively. The recognition accuracy of the improved
Alexnet algorithm is 18.76% higher than that of the unimproved one.
After multiple iterations, it is verified that the recognition accuracy
and loss function are better than those of the unimproved Alexnet
algorithm. It shows that the identification system proposed can better
monitor and identify vegetable pests and diseases, which is beneficial
to integrated management.