The identification of maize leaf diseases will meet great challenges because of the difficulties in extracting lesion features from the constant-changing environment, uneven illumination reflection of the incident light source and many other factors. In this paper, a novel maize leaf disease recognition method is proposed. In this method, we first designed a maize leaf feature enhancement framework with the capability of enhancing the features of maize under the complex environment. Then a novel neural network is designed based on backbone Alexnet architecture, named DMS-Robust Alexnet. In the DMS-Robust Alexnet, dilated convolution and multi-scale convolution are combined to improve the capability of feature extraction. Batch normalization is performed to prevent network over-fitting while enhancing the robustness of the model. PRelu activation function and Adabound optimizer are employed to improve both convergence and accuracy. In experiments, it is validated from different perspectives that the maize leaf disease feature enhancement algorithm is conducive to improving the capability of the DMS-Robust Alexnet identification. Our method demonstrates strong robustness for maize disease images collected in the natural environment, providing a reference for the intelligent diagnosis of other plant leaf diseases.

Aiming at the problems of noise, background interference and low detection in peach disease image, this paper proposes a detection method of peach disease based on the asymptotic non-local means (ANLM) image algorithm and the fusion of parallel convolution neural network (PCNN) and extreme learning machine(ELM) optimized by linear particle swarm optimization(IPSO). Firstly, the method uses the ANLM image denoising algorithm to reduce the interference of the complex background in the image, then uses the parallel convolution neural network proposed by this paper to identify the characteristics of peach disease, uses the improved elu activation function instead of the conventional ReLu activation function, and uses the linear particle swarm optimized ELM (IPELM) proposed by this paper in the last layer instead of the traditional softmax layer, the update method is improved from two aspects to improve the convergence speed and accuracy of the network effectively. The results of 25513 images showed that the highest detection accuracy of brown rot, black spot, anthracnose, scab and normal peach were 89.02, 90.56, 85.37, 86.70 and 89.91 percent respectively, which indicated that this method was an effective method for peach disease detection.

To solve the complex background problems (e.g. Noise interference, object overlap, and different illumination) that affect the classification performance on plant leaf images, this paper proposes an instance segmentation and classification method for plant leaf images based on IFPN SNMS CFFI-Mask R-CNN (ISC-MRCNN) and ACPSOSVM-Dual Channels Convolutional Neural Network (APS-DCCNN). To obtain the foreground of plant leaf images, the lateral connection structure of the feature map pyramid in ISC-MRCNN fuses the feature maps of different depths, so that the network learns more detailed features. Then, the Soft Non-Maximum Suppression Algorithm is employed to improve the detection performance of overlapping objects. Next, the pooling method of integrating the continuous function can reduce the precision loss during the alignment of the mapping between the feature map and the original image. Finally, by constructing a mask filter layer, complex backgrounds are masked. To distinguish the similarity between plant leaf images, APS-DCCNN is used to classify the foreground images. In this process, the Support Vector Machine is used to replace softmax and then an Adaptive Chaotic Particle Swarm Algorithm is employed to optimize it. The experimental results show that compared with Mask R-CNN, the average precision of ISC-MRCNN has increased by 1.89% under different thresholds. The proposed method is suitable for the object detection and instance segmentation problems with complex background. Besides, compared with traditional CNN, the average precision of the classification results obtained by APS-DCCNN has improved by 1.59%. This has shown that the proposed method is suitable for the classification of plant leaves.