Abstract:

The outbreak of crop pests and diseases often means large scales of production and quality decline which can lead to irrecoverable economic losses. The traditional disease and pest identification approach is slow, subjective and not accurate, not suitable for the current agricultural production. The image processing based crop disease and pest identification is fast, accurate and realtime, and it can assist farmers to timely adopt effective preventative methods. This thesis describes the study status and progress of image processing technology’s application on crop disease and pest identification in image segmentation, feature extraction and classification, and outlooks into the future trend and direction of further studies.

Key words: image processing;diseases and pests;segmentation;feature extraction;classification and recognition