Deep Transfer Learning based Acoustic Detection of Rice Weevils, Sitophilus Oryzae (L.) in Stored Grains

 
 
 
  • Abstract
  • Keywords
  • References
  • PDF
  • Abstract


    The presence of rice weevils is causing degradation of rice quantity and quality during storage. Classifying rice grade is critical since rice weevils are not easily detected. This study used deep transfer learning on spectrogram images of sounds to recognize the presence or absence of rice weevils in a sound clip. There are 1000 audio files with rice weevil presence and 1000 audio files with the absence of rice weevils in the dataset, each having a duration of 5 seconds. Random environments and random number and age of insects were considered to have models that are less dependent on the environment setting. The dataset was preprocessed to generate the spectrogram image of each audio clip. Features of those images were extracted to train some pre-trained Keras models on the dataset. In the dataset, 1400 images were used for training and 600 were used for testing. Each among the models Xception, ResNet50, InceptionResNetV2, and MobileNet obtained a rank-1 accuracy of 99.17% while VGG16, VGG19, and InceptionV3 all got a rank-1 accuracy of 99.33%. The average precision, average recall, and average F1 score in each trained model are all 99%. These account for the effectiveness of using deep transfer learning on spectrogram images of audio recordings in the detection of rice weevils in stored grains. This is also the first study that used deep transfer learning on spectrogram images in the acoustic detection of rice weevils.

     

     


  • Keywords


    acoustic detection, deep transfer learning, librosa, Keras, Sitophilus oryzae (L.).

  • References


      [1] Neethirajan S, Karunakaran C, Jayas DS & White NDG (2007), Detection techniques for stored-product insects in grain. Food Control 18(2), 157-162.

      [2] Burks CS, Yasin M, El-Shafie HA & Wakil W (2015), Pests of stored dates. Sustainable pest management in date palm: current status and emerging challenges, 237-286.

      [3] Eliopoulos PA, Potamitis I & Kontodimas DC (2016), Estimation of population density of stored grain pests via bioacoustic detection. Crop Protection 85, 71-78.

      [4] Eliopoulos PA, Potamitis I, Kontodimas DC & Givropoulou EG (2015), Detection of adult beetles inside the stored wheat mass based on their acoustic emissions. Journal of economic entomology 108(6), 2808-2814.

      [5] Fleurat-Lessard F, Tomasini B, Kostine L & Fuzeau B (2006), Acoustic detection and automatic identification of insect stages activity in grain bulks by noise spectra processing through classification algorithms. Conference Working on Stored Product Protection.

      [6] Le-Qing Z (2011). Insect sound recognition based on mfcc and pnn. Multimedia and Signal Processing (CMSP), 2011 International Conference 2, 42-46.

      [7] Yazgaç BG, Kırcı M & Kıvan M. (2016), Detection of sunn pests using sound signal processing methods. Agro-Geoinformatics (Agro-Geoinformatics), 2016 Fifth International Conference, 1-6.

      [8] Noda JJ, Travieso CM, Sánchez-Rodríguez D, Dutta MK & Singh A (2016), Using bioacoustic signals and support vector machine for automatic classification of insects. Signal Processing and Integrated Networks (SPIN), 2016 3rd International Conference, 656-659.

      [9] Flynn T, Salloum H, Hull-Sanders H, Sedunov A, Sedunov N, Sinelnikov Y, Sutin A & Masters D (2016), Acoustic methods of invasive species detection in agriculture shipments. Technologies for Homeland Security (HST), 2016 IEEE Symposium, 1-5.

      [10] Pan W, Kong X, Xu J & Pan W (2016), Measurement and analysis system of vibration for the detection of insect acoustic signals. Electromagnetic Compatibility (APEMC), 2016 Asia-Pacific International Symposium 1, 1090-1092.

      [11] Peso Parada P & Cardenal-López A (2014), Using Gaussian mixture models to detect and classify dolphin whistles and pulses. The Journal of the Acoustical Society of America 135(6), 3371-3380.

      [12] Gupta D, Jain S, Shaikh F & Singh G (2017), Transfer learning & the art of using pre-trained models in deep learning. Analytics Vidhya, available online: https://www.analyticsvidhya.com/blog/2017/06/transfer-learning-the-art-of-fine-tuning-a-pre-trained-model/, last visit: 07.07.2018.

      [13] Chollet F, Deep learning with python, Manning Publications Co. (2017).

      [14] Gulli A & Pal S, Deep Learning with Keras, Packt Publishing Ltd., (2017).

      [15] Kieffer B, Babaie M, Kalra S & Tizhoosh HR (2017), Convolutional neural networks for histopathology image classification: training vs. using pre-trained networks. 2017 Seventh International Conference on Image Processing Theory, Tools and Applications (IPTA), 1-6.

      [16] Mankin RW & Fisher JR (2002), Acoustic detection of black vine weevil, Otiorhynchus sulcatus (Fabricius)(Coleoptera: Curculionidae) larval infestations in nursery containers. Journal of Environmental Horticulture, 20(3), 166-170.

      [17] Piczak KJ (2015), ESC: Dataset for environmental sound classification. Proceedings of the 23rd ACM international conference on Multimedia, 1015-1018.

      [18] McFee B, Raffel C, Liang D, Ellis DP, McVicar M, Battenberg E & Nieto O (2015), librosa: Audio and music signal analysis in python. Proceedings of the 14th python in science conference, 18-25.

      [19] Oliva SL, Palmieri A, Invidia L, Patrono L & Rametta P (2018), Rapid Prototyping of a Star Topology Network based on Bluetooth Low Energy Technology. 2018 3rd International Conference on Smart and Sustainable Technologies (SpliTech), 1-6.

      [20] Cortez D, Molina C, Abante M, Santos M, and Sarmiento M (2018), Nullpest: a mobile application of agricultural pest locator using sonar sensor set-up. International Journal of Engineering & Technology 7(3.29), 107-109.


 

View

Download

Article ID: 21785
 
DOI: 10.14419/ijet.v7i4.16.21785




Copyright © 2012-2015 Science Publishing Corporation Inc. All rights reserved.