A rapid rice blast detection and identification method based on crop disease spores' diffraction fingerprint texture

Area of Science:

• Agricultural Science
• Plant Pathology
• Biotechnology

Background:

• Rice blast fungus poses a significant global threat to rice production, particularly in China.
• Early symptoms are subtle, and rapid spread makes manual identification challenging and inefficient.
• Current methods for early rice blast detection are time-consuming and labor-intensive.

Purpose of the Study:

• To develop a rapid and accurate method for early identification of rice blast.
• To investigate the utility of diffraction fingerprint texture for rice blast spore identification.
• To compare the proposed method with traditional manual identification and machine recognition techniques.

Main Methods:

• Utilized light field and texture features of diffraction images of rice blast spores.
• Developed an identification method based on crop disease spores' diffraction fingerprint texture.
• Compared the proposed method against manual identification via microscope and machine recognition.

Main Results:

• The diffraction recognition method achieved a test accuracy of 97.18%.
• Identification accuracy using light diffraction characteristics was over 0.3% higher than manual microscopy.
• The diffraction recognition method, after neural network training, was over 90% faster than manual methods, completing detection in seconds.

Conclusions:

• The proposed diffraction fingerprint texture method provides a rapid and advantageous alternative to manual microscopy for rice blast detection.
• This technique is suitable for recognizing rice blast in agricultural research settings.
• The study highlights the potential of optical properties of spores for disease identification.