Deep Learning-Based Segmentation and Quantification of Cucumber Powdery Mildew Using Convolutional Neural Network

Abstract

Powdery mildew is a common disease in plants, and it is also one of the main diseases in the middle and final stages of cucumber (Cucumis sativus). Powdery mildew on plant leaves affects the photosynthesis, which may reduce the plant yield. Therefore, it is of great significance to automatically identify powdery mildew. Currently, most image-based models commonly regard the powdery mildew identification problem as a dichotomy case, yielding a true or false classification assertion. However, quantitative assessment of disease resistance traits plays an important role in the screening of breeders for plant varieties. Therefore, there is an urgent need to exploit the extent to which leaves are infected which can be obtained by the area of diseases regions. In order to tackle these challenges, we propose a semantic segmentation model based on convolutional neural networks (CNN) to segment the powdery mildew on cucumber leaf images at pixel level, achieving an average pixel accuracy of 96.08%, intersection over union of 72.11% and Dice accuracy of 83.45% on twenty test samples. This outperforms the existing segmentation methods, K-means, Random forest, and GBDT methods. In conclusion, the proposed model is capable of segmenting the powdery mildew on cucumber leaves at pixel level, which makes a valuable tool for cucumber breeders to assess the severity of powdery mildew.

Aerial Visual Perception in Smart Farming: Field Study of Wheat Yellow Rust Monitoring

“…Various types of sensors have been investigated for disease monitoring in the literature: from low-cost RGB visual [9] to high-cost hyperspectral camera [6] and from ground proximity sensing [10] to aircraft (or even satellite) remote sensing [7].…”

“…The encoder-decoder cascaded CNN, SegNet, is also applied in [12] for weed mapping by using multispectral image. Very recently, the state-of-the-art U-Net is applied in [10] for leaf level disease segmentation of cucumber leaf with promising performance. U-Net and mask R-CNN [16] are compared [13] for tree canopy segmentation by using UAV RGB image at 30m.…”

Efficient and Automated Herbs Classification Approach Based on Shape and Texture Features using Deep Learning

“…Currently, several attempts have been made to use Deep Learning Neural Networks (DLNN) for the classification of herbs [86][87][88][89][90][91][92][93][94][95][96][97][98][99], and plants diseases [100][101][102][103][104][105][106][107]. The motivation is to attract attention to such methods after the state-of-the-art processing of natural images [84,85].…”

Artificial Neural Network Modeling of Greenhouse Tomato Yield and Aerial Dry Matter

“…ANNs allow us to develop models based on the intrinsic relations among the variables, without prior knowledge of their functional relationships [9]. Soft computing for ANN techniques has been widely used to develop models to predict different crop indicators, such as growth, yield, and other biophysical processes, and also because of the commercial importance of tomato [10][11][12][13][14][15][16][17][18][19][20][21][22][23] and other vegetables, such as lettuce [24][25][26][27][28][29][30], pepper [31][32][33][34], cucumber [35][36][37][38], wheat [39][40][41][42][43][44][45], rice [46][47][48], oat [49], maize [50,51], corn [52][53][54], corn and soybean [55], soybean…”