Performance Analysis of Optimizers for Plant Disease Classification with Convolutional Neural Networks

Labhsetwar, Shreyas Rajesh; Haridas, Soumya; Panmand, Riyali; Deshpande, Rutuja; Kolte, Piyush Arvind; Pati, Sandhya

doi:10.1109/icnte51185.2021.9487698

Cited by 6 publications

(9 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An optimizer’s goal is to minimize the loss function by updating the weights and biases in each iteration of training using the training rate. The most used optimizer is stochastic gradient descent (SGD); however, this optimizer needs to be tuned throughout the training process [ 76 ]. To mitigate this problem, the functionality of the learning rate has been modified according to different aspects and named Adagrad, Adadelta, RMS-Prop, and ADAM [ 45 , 77 ].…”

Section: Methodsmentioning

confidence: 99%

Autonomous Image-Based Corrosion Detection in Steel Structures Using Deep Learning

Das,

Dorafshan,

Kaabouch

2024

Sensors

View full text Add to dashboard Cite

Steel structures are susceptible to corrosion due to their exposure to the environment. Currently used non-destructive techniques require inspector involvement. Inaccessibility of the defective part may lead to unnoticed corrosion, allowing the corrosion to propagate and cause catastrophic structural failure over time. Autonomous corrosion detection is essential for mitigating these problems. This study investigated the effect of the type of encoder–decoder neural network and the training strategy that works the best to automate the segmentation of corroded pixels in visual images. Models using pre-trained DesnseNet121 and EfficientNetB7 backbones yielded 96.78% and 98.5% average pixel-level accuracy, respectively. Deeper EffiecientNetB7 performed the worst, with only 33% true-positive values, which was 58% less than ResNet34 and the original UNet. ResNet 34 successfully classified the corroded pixels, with 2.98% false positives, whereas the original UNet predicted 8.24% of the non-corroded pixels as corroded when tested on a specific set of images exclusive to the investigated training dataset. Deep networks were found to be better for transfer learning than full training, and a smaller dataset could be one of the reasons for performance degradation. Both fully trained conventional UNet and ResNet34 models were tested on some external images of different steel structures with different colors and types of corrosion, with the ResNet 34 backbone outperforming conventional UNet.

show abstract

Section: Methodsmentioning

confidence: 99%

Autonomous Image-Based Corrosion Detection in Steel Structures Using Deep Learning

Das,

Dorafshan,

Kaabouch

2024

Sensors

View full text Add to dashboard Cite

show abstract

“…Disturbing backgrounds scenes and objects, such as soil and other biomass, create problems in target image annotation for visible images in automated disease and weed detection and [30,31]. Noise, blurring, brightness, and contrast issues can degrade the image quality, where image noise results from the interaction of natural light and camera mechanics [32]. Due to the speed of UAVs, captured images could be impacted by motion blur and excessive brightness, posing a significant challenge for classification and object detection [33].…”

Section: Challenges In Agricultural Image Datasetsmentioning

confidence: 99%

Pansharpening Low-Altitude Multispectral Images of Potato Plants Using a Generative Adversarial Network

Modak,

Heil,

Stein

2024

Remote Sensing

View full text Add to dashboard Cite

Image preprocessing and fusion are commonly used for enhancing remote-sensing images, but the resulting images often lack useful spatial features. As the majority of research on image fusion has concentrated on the satellite domain, the image-fusion task for Unmanned Aerial Vehicle (UAV) images has received minimal attention. This study investigated an image-improvement strategy by integrating image preprocessing and fusion tasks for UAV images. The goal is to improve spatial details and avoid color distortion in fused images. Techniques such as image denoising, sharpening, and Contrast Limited Adaptive Histogram Equalization (CLAHE) were used in the preprocessing step. The unsharp mask algorithm was used for image sharpening. Wiener and total variation denoising methods were used for image denoising. The image-fusion process was conducted in two steps: (1) fusing the spectral bands into one multispectral image and (2) pansharpening the panchromatic and multispectral images using the PanColorGAN model. The effectiveness of the proposed approach was evaluated using quantitative and qualitative assessment techniques, including no-reference image quality assessment (NR-IQA) metrics. In this experiment, the unsharp mask algorithm noticeably improved the spatial details of the pansharpened images. No preprocessing algorithm dramatically improved the color quality of the enhanced images. The proposed fusion approach improved the images without importing unnecessary blurring and color distortion issues.

show abstract

“…CNN has been used to detect fifteen types of diseases in three major crops (tomato, potato, and pepper). They achieved a maximum detection rate of 98% using different optimizer functions in CNN [ 23 ]. According to Guan et al, four types of DNNs (VGG19, VGG16, Inception-v3, and ResNet50) have been developed to classify apple plants’ disease severity levels.…”

Section: State-of-the-art Work Related To Leaf Disease Detectionmentioning

confidence: 99%

Brinjal leaf diseases detection based on discrete Shearlet transform and Deep Convolutional Neural Network

Abisha¹,

Mutawa

Murugappan

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

Different diseases are observed in vegetables, fruits, cereals, and commercial crops by farmers and agricultural experts. Nonetheless, this evaluation process is time-consuming, and initial symptoms are primarily visible at microscopic levels, limiting the possibility of an accurate diagnosis. This paper proposes an innovative method for identifying and classifying infected brinjal leaves using Deep Convolutional Neural Networks (DCNN) and Radial Basis Feed Forward Neural Networks (RBFNN). We collected 1100 images of brinjal leaf disease that were caused by five different species (Pseudomonas solanacearum, Cercospora solani, Alternaria melongenea, Pythium aphanidermatum, and Tobacco Mosaic Virus) and 400 images of healthy leaves from India’s agricultural form. First, the original plant leaf is preprocessed by a Gaussian filter to reduce the noise and improve the quality of the image through image enhancement. A segmentation method based on expectation and maximization (EM) is then utilized to segment the leaf’s-diseased regions. Next, the discrete Shearlet transform is used to extract the main features of the images such as texture, color, and structure, which are then merged to produce vectors. Lastly, DCNN and RBFNN are used to classify brinjal leaves based on their disease types. The DCNN achieved a mean accuracy of 93.30% (with fusion) and 76.70% (without fusion) compared to the RBFNN (82%—without fusion, 87%—with fusion) in classifying leaf diseases.

show abstract

Performance Analysis of Optimizers for Plant Disease Classification with Convolutional Neural Networks

Cited by 6 publications

References 21 publications

Autonomous Image-Based Corrosion Detection in Steel Structures Using Deep Learning

Autonomous Image-Based Corrosion Detection in Steel Structures Using Deep Learning

Pansharpening Low-Altitude Multispectral Images of Potato Plants Using a Generative Adversarial Network

Brinjal leaf diseases detection based on discrete Shearlet transform and Deep Convolutional Neural Network

Contact Info

Product

Resources

About