Deep convolutional neural networks for rice grain yield estimation at the ripening stage using UAV-based remotely sensed images

Skin cancer is one of the top three perilous types of cancer caused by damaged DNA that can cause death. This damaged DNA begins cells to grow uncontrollably and nowadays it is getting increased speedily. There exist some researches for the computerized analysis of malignancy in skin lesion images. However, analysis of these images is very challenging having some troublesome factors like light reflections from the skin surface, variations in color illumination, different shapes, and sizes of the lesions. As a result, evidential automatic recognition of skin cancer is valuable to build up the accuracy and proficiency of pathologists in the early stages. In this paper, we propose a deep convolutional neural network (DCNN) model based on deep learning approach for the accurate classification between benign and malignant skin lesions. In preprocessing we firstly, apply filter or kernel to remove noise and artifacts; secondly, normalize the input images and extract features that help for accurate classification; and finally, data augmentation increases the number of images that improves the accuracy of classification rate. To evaluate the performance of our proposed, DCNN model is compared with some transfer learning models such as AlexNet, ResNet, VGG-16, DenseNet, MobileNet, etc. The model is evaluated on the HAM10000 dataset and ultimately we obtained the highest 93.16% of training and 91.93% of testing accuracy respectively. The final outcomes of our proposed DCNN model define it as more reliable and robust when compared with existing transfer learning models.

An enhanced technique of skin cancer classification using deep convolutional neural network with transfer learning models

Alcohol use disorder (AUD) is an important brain disease. It alters the brain structure. Recently, scholars tend to use computer vision based techniques to detect AUD. We collected 235 subjects, 114 alcoholic and 121 non-alcoholic. Among the 235 image, 100 images were used as training set, and data augmentation method was used. The rest 135 images were used as test set. Further, we chose the latest powerful technique-convolutional neural network (CNN) based on convolutional layer, rectified linear unit layer, pooling layer, fully connected layer, and softmax layer. We also compared three different pooling techniques: max pooling, average pooling, and stochastic pooling. The results showed that our method achieved a sensitivity of 96.88%, a specificity of 97.18%, and an accuracy of 97.04%. Our method was better than three state-of-the-art approaches. Besides, stochastic pooling performed better than other max pooling and average pooling. We validated CNN with five convolution layers and two fully connected layers performed the best. The GPU yielded a 149× acceleration in training and a 166× acceleration in test, compared to CPU.

Alcoholism Detection by Data Augmentation and Convolutional Neural Network with Stochastic Pooling

For hyperspectral remote sensing image (HSI) classification, the learning process of deep neural networks has been progressively advanced in depth, but the fine features are often largely lost or even disappear in the process of depth transfer. With the increase in feature aggregation and connectivity, the complexity of the network and the training parameters increases greatly, requiring more training time. This paper proposed a multi-scale dense network (MSDN) for HSI classification that made full use of different scale information in the network structure and combined scale information throughout the network. It implemented feature extraction of HSIs in two dimensions, including the features at fine and coarse levels. In the horizontal direction, it considered the deep extraction of HSI features, and the 3-D dense connection structure was used for aggregating features at different levels. In the vertical direction, scale information was considered, and three-scale feature maps at low, middle, and high levels were generated based on the first layer of the network. The MSDN used stride convolution for downsampling and combined feature information at different scale levels. The MSDN extracted features along the diagonal line. The network implemented the reconstruction of deep feature extraction and multi-scale fusion for HSI classification. The MSDN model performed well on representative HSI datasets, namely, the Indian Pines, Pavia University, Salinas, Botswana, and Kennedy Space Center datasets. It improved the training speed and accuracy for HSI classification and especially improved the convergence speed, which effectively saved computing resources and had high stability.

Multi-Scale Dense Networks for Hyperspectral Remote Sensing Image Classification

Understanding land surface temperature impact factors based on local climate zones

The rapid development of high spatial resolution (HSR) remote sensing imagery techniques not only provide a considerable amount of datasets for scene classification tasks but also request an appropriate scene classification choice when facing with finite labeled samples. AlexNet, as a relatively simple convolutional neural network (CNN) architecture, has obtained great success in scene classification tasks and has been proven to be an excellent foundational hierarchical and automatic scene classification technique. However, current HSR remote sensing imagery scene classification datasets always have the characteristics of small quantities and simple categories, where the limited annotated labeling samples easily cause non-convergence. For HSR remote sensing imagery, multi-scale information of the same scenes can represent the scene semantics to a certain extent but lacks an efficient fusion expression manner. Meanwhile, the current pre-trained AlexNet architecture lacks a kind of appropriate supervision for enhancing the performance of this model, which easily causes overfitting. In this paper, an improved pre-trained AlexNet architecture named pre-trained AlexNet-SPP-SS has been proposed, which incorporates the scale pooling—spatial pyramid pooling (SPP) and side supervision (SS) to improve the above two situations. Extensive experimental results conducted on the UC Merced dataset and the Google Image dataset of SIRI-WHU have demonstrated that the proposed pre-trained AlexNet-SPP-SS model is superior to the original AlexNet architecture as well as the traditional scene classification methods.

Pre-Trained AlexNet Architecture with Pyramid Pooling and Supervision for High Spatial Resolution Remote Sensing Image Scene Classification

The formation of black-odor water in urban rivers has a long history. It not only seriously affects the image of the city, but also easily breeds germs and damages the urban habitat. The prevention and treatment of urban black-odor water have long been important topics nationwide. “Action Plan for Prevention and Control of Water Pollution” issued by the State Council shows Chinese government’s high attention to this issue. However, treatment and monitoring are inextricably linked. There are few studies on the large-scale monitoring of black-odor water, especially the cases of using unmanned aerial vehicle (UAV) to efficiently and accurately monitor the spatial distribution of urban river pollution. Therefore, in order to get rid of the limitations of traditional ground sampling to evaluate the point source pollution of rivers, the UAV-borne hyperspectral imagery was applied in this paper. It is hoped to grasp the pollution status of the entire river as soon as possible from the surface. However, the retrieval of multiple water quality parameters will lead to cumulative errors, so the Nemerow comprehensive pollution index (NCPI) is introduced to characterize the pollution level of urban water. In the paper, the retrieval results of six regression models including gradient boosting decision tree regression (GBDTR) were compared, trying to find a regression model for the retrieval NCPI in the current scenario. In the first study area, the retrieval accuracy of the training dataset (adjusted_R2 = 0.978), and test dataset (adjusted_R2 = 0.974) was higher than that of the other regression models. Although the retrieval effect of random forest is similar to that of GBDTR in both training accuracy and image inversion, it is more computationally expensive. Finally, the spatial distribution graphs of NCPI and its technical feasibility in monitoring pollution sources were investigated, in combination with field observations.

Monitoring of Urban Black-Odor Water Based on Nemerow Index and Gradient Boosting Decision Tree Regression Using UAV-Borne Hyperspectral Imagery

Soil organic matter (SOM) refers to all carbon-containing organic matter in soil and is oneof the most important indicators of soil fertility. The hyperspectral inversion analysis of SOMtraditionally relies on laboratory chemical testing methods, which have the disadvantages of beinginefficient and time-consuming. In this study, 69 soil samples were collected from the Honghufarmland area and a mining area in northwest China. After pretreatment, 10 spectral indicators wereobtained. Ridge regression, kernel ridge regression, Bayesian ridge regression, and AdaBoostalgorithms were then used to construct the SOM hyperspectral inversion model based on thecharacteristic bands, and the accuracy of the models was compared. The results showed that theAdaBoost algorithm based on a grid search had the best accuracy in the different regions. For themining area in northwest China [...].

https://www.mdpi.com/1424-8220/20/10/2777/pdf

Hyperspectral Inversion of Soil Organic Matter Content Based on a Combined Spectral Index Model.

In this study, in order to solve the difficulty of the inversion of soil arsenic (As) content using laboratory and field reflectance spectroscopy, we examined the transferability of the prediction method. Sixty-three soil samples from the Daye city area of the Jianghan Plain region of China were taken and studied in this research. The characteristic wavelengths of soil As content were then extracted from the full bands based on iteratively retaining informative variables (IRIV) coupled with Spearman's rank correlation analysis (SCA). Firstly, the IRIV algorithm was used to roughly select the original spectral data. Gaussian filtering (GF), first derivative (FD) filtering, and gaussian filtering again (GFA) pretreatments were then used to improve the correlation between the spectra and soil As content. A subset with absolute correlation values greater than 0.6 was then retained as the optimal subset after each pretreatment. Finally, partial least squares regression (PLSR), Bayesian ridge regression (BRR), ridge regression (RR), kernel ridge regression (KRR), support vector machine regression (SVMR), eXtreme gradient boosting (XGBoost) regression, and random forest regression (RFR) models were used to estimate the soil As values using the different characteristic variables. The results showed that, compared with the traditional method based on IRIV, using the characteristic bands selected by the IRIV-SCA method can effectively improve the prediction accuracy of the models. For the laboratory spectra experiment stage, the six most representative characteristic bands were selected. The performance of IRIV-SCA-SVMR was found to be the best, with the coefficient of determination (R2), root-mean-square error (RMSE), and mean absolute error (MAE) in the validation set being 0.97, 0.22, and 0.11, respectively. For the field spectra experiment stage, the 12 most representative characteristic bands were selected. The performance of IRIV-SCA-XGBoost was found to be the best, with the R2, RMSE, and MAE in the validation set being 0.83, 0.35, and 0.29, respectively. The accuracy and stability of the inversion of soil As content are significantly improved by the use of the proposed method, and the method could be used to provide accurate data for decision support for the treatment and recovery of As pollution over a large area.

Estimation of Arsenic Content in Soil Based on Laboratory and Field Reflectance Spectroscopy.

Hyperspectral imagery has been widely used in precision agriculture due to its rich spectral characteristics. With the rapid development of remote sensing technology, the airborne hyperspectral imagery shows detailed spatial information and temporal flexibility, which open a new way to accurate agricultural monitoring. To extract crop types from the airborne hyperspectral images, we propose a fine classification method based on multi-feature fusion and deep learning. In this research, the morphological profiles, GLCM texture and endmember abundance features are leveraged to exploit the spatial information of the hyperspectral imagery. Then, the multiple spatial information is fused with the original spectral information to generate classification result by using the deep neural network with conditional random field (DNN+CRF) model. Specifically, the deep neural network (DNN) is a deep recognition model which can extract depth features and mine the potential information of data. As a discriminant model, conditional random field (CRF) considers both spatial and contextual information to reduce the misclassification noises while keeping the object boundaries. Moreover, three multiple feature fusion approaches, namely feature stacking, decision fusion and probability fusion, are taken into account. In the experiments, two airborne hyperspectral remote sensing datasets (Honghu dataset and Xiong’an dataset) are used. The experimental results show that the classification performance of the proposed method is satisfactory, where the salt and pepper noise is decreased, and the boundary of the ground object is preserved.

https://www.mdpi.com/2072-4292/13/15/2917/pdf

Liqin Cao

Papers

Pre-Trained AlexNet Architecture with Pyramid Pooling and Supervision for High Spatial Resolution Remote Sensing Image Scene Classification

Monitoring of Urban Black-Odor Water Based on Nemerow Index and Gradient Boosting Decision Tree Regression Using UAV-Borne Hyperspectral Imagery

Hyperspectral Inversion of Soil Organic Matter Content Based on a Combined Spectral Index Model.

Estimation of Arsenic Content in Soil Based on Laboratory and Field Reflectance Spectroscopy.

Crops Fine Classification in Airborne Hyperspectral Imagery Based on Multi-Feature Fusion and Deep Learning