SPAD monitoring of saline vegetation based on Gaussian mixture model and UAV hyperspectral image feature classification

Zhu, Chuanmei; Ding, Jianli; Zhang, Zipeng; Wang, Jinjie; Wang, Zheng; Chen, Xiangyue; Wang, Jingzhe

doi:10.1016/j.compag.2022.107236

Cited by 22 publications

(13 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It was found through literature review that the red, yellow and blue edge spectral parameters have been frequently used in crop quality monitoring and forecast (Curran 1989, Lamb et al 2002, Olivares Díaz et al 2019, Zhu et al 2022). In the present study, the following spectral parameters were screened and chosen useful ones to build the prediction model: field canopy spectra, firstorder derivative spectra of the field canopy, five vegetation indices (NDVI, RVI, EVI, DVI, and TVI), three-edge parameters (red, blue, and yellow edges), red valley position, and green peak position (Table 1).…”

Section: Methodsmentioning

confidence: 99%

Hyperspectral multivariate linear prediction model of tobacco (nicotiana tabacum L.) Leaf nitrogen content

Guo,

Li,

et al. 2023

Bangladesh J. Bot.

View full text Add to dashboard Cite

In order to accurately and effectively obtain the nitrogen content of tobacco leaves during the whole growth period, in the present study the field canopy spectrum of the three critical periods of tobacco rosette stage, vigorous growth stage and topping stage were used. The correlation analysis of field canopy spectrum, first derivative spectrum, hyperspectral parameters and vegetation index with the nitrogen content of tobacco leaves was carried out one by one, and the prediction model was established by multiple linear regression using the variables with the best correlation coefficient. Results showed that the first derivative spectrum, EVI II and green peak position had strong correlation, which is suitable for introducing multivariate equations as independent variables. Finally, the modeling determination coefficient (R2) was 0.66, RMSE was 0.40, and MAPE was 11%. The validation results showed that R2 was 0.73, RMSE was 0.38, and MAPE was 8.33%, which proved that this model could accurately predict the nitrogen content of tobacco leaves and could meet the requirements of large-scale statistical monitoring of tobacco quality indicators in the field. Bangladesh J. Bot. 52(2): 575-584, 2023 (June) Special

show abstract

Section: Methodsmentioning

confidence: 99%

Hyperspectral multivariate linear prediction model of tobacco (nicotiana tabacum L.) Leaf nitrogen content

Guo,

Li,

et al. 2023

Bangladesh J. Bot.

View full text Add to dashboard Cite

show abstract

“…This study tackled this issue by automatically splitting image pixels with different spectral characteristics into spectrally more homogenous subgroups via unsupervised clustering by utilizing the Gaussian mixture model (GMM) [ 34 , 35 ]. The GMM is based on the characterization of a heterogenous input data distribution with a linear mixture of unimodal Gaussian distributions and has been used in many different HSI applications, such as hyperspectral image segmentation [ 36 , 37 ], the monitoring of saline vegetation [ 38 ], and anomaly detection [ 39 ]. Given the input data vectors in an n -dimensional space and the number of clusters K , the GMM estimated the distribution of a data vector x with K unimodal Gaussian distributions that were linearly mixed in the following equation:

where

is the k th normal distribution with a mean of

and a covariance matrix of

, and

is the weight of the k th Gaussian distribution.…”

Section: Methodsmentioning

confidence: 99%

Semisupervised Deep Learning for the Detection of Foreign Materials on Poultry Meat with Near-Infrared Hyperspectral Imaging

Campos

Yoon

Chung

et al. 2023

Sensors

View full text Add to dashboard Cite

A novel semisupervised hyperspectral imaging technique was developed to detect foreign materials (FMs) on raw poultry meat. Combining hyperspectral imaging and deep learning has shown promise in identifying food safety and quality attributes. However, the challenge lies in acquiring a large amount of accurately annotated/labeled data for model training. This paper proposes a novel semisupervised hyperspectral deep learning model based on a generative adversarial network, utilizing an improved 1D U-Net as its discriminator, to detect FMs on raw chicken breast fillets. The model was trained by using approximately 879,000 spectral responses from hyperspectral images of clean chicken breast fillets in the near-infrared wavelength range of 1000–1700 nm. Testing involved 30 different types of FMs commonly found in processing plants, prepared in two nominal sizes: 2 × 2 mm2 and 5 × 5 mm2. The FM-detection technique achieved impressive results at both the spectral pixel level and the foreign material object level. At the spectral pixel level, the model achieved a precision of 100%, a recall of over 93%, an F1 score of 96.8%, and a balanced accuracy of 96.9%. When combining the rich 1D spectral data with 2D spatial information, the FM-detection accuracy at the object level reached 96.5%. In summary, the impressive results obtained through this study demonstrate its effectiveness at accurately identifying and localizing FMs. Furthermore, the technique’s potential for generalization and application to other agriculture and food-related domains highlights its broader significance.

show abstract

“…The RF algorithm, proposed by Breiman (2001) , is a popular ensemble learning algorithm in classification, prediction, and feature selection ( Breiman, 2001 ). When using the RF algorithm for classification, the final label of the input sample is determined by voting for each decision tree in the random forest ( Guo et al, 2011 ; Zhu et al, 2022b ). Random resampling and node random splitting techniques are used to train the RF model ( Gislason et al, 2006 ).…”

Section: Methodsmentioning

confidence: 99%

Detection of wheat Fusarium head blight using UAV-based spectral and image feature fusion

Zhang

Huang

et al. 2022

Front. Plant Sci.

View full text Add to dashboard Cite

Infection caused by Fusarium head blight (FHB) has severely damaged the quality and yield of wheat in China and threatened the health of humans and livestock. Inaccurate disease detection increases the use cost of pesticide and pollutes farmland, highlighting the need for FHB detection in wheat fields. The combination of spectral and spatial information provided by image analysis facilitates the detection of infection-related damage in crops. In this study, an effective detection method for wheat FHB based on unmanned aerial vehicle (UAV) hyperspectral images was explored by fusing spectral features and image features. Spectral features mainly refer to band features, and image features mainly include texture and color features. Our aim was to explain all aspects of wheat infection through multi-class feature fusion and to find the best FHB detection method for field wheat combining current advanced algorithms. We first evaluated the quality of the two acquired UAV images and eliminated the excessively noisy bands in the images. Then, the spectral features, texture features, and color features in the images were extracted. The random forest (RF) algorithm was used to optimize features, and the importance value of the features determined whether the features were retained. Feature combinations included spectral features, spectral and texture features fusion, and the fusion of spectral, texture, and color features to combine support vector machine, RF, and back propagation neural network in constructing wheat FHB detection models. The results showed that the model based on the fusion of spectral, texture, and color features using the RF algorithm achieved the best performance, with a prediction accuracy of 85%. The method proposed in this study may provide an effective way of FHB detection in field wheat.

show abstract

SPAD monitoring of saline vegetation based on Gaussian mixture model and UAV hyperspectral image feature classification

Cited by 22 publications

References 39 publications

Hyperspectral multivariate linear prediction model of tobacco (nicotiana tabacum L.) Leaf nitrogen content

Hyperspectral multivariate linear prediction model of tobacco (nicotiana tabacum L.) Leaf nitrogen content

Semisupervised Deep Learning for the Detection of Foreign Materials on Poultry Meat with Near-Infrared Hyperspectral Imaging

Detection of wheat Fusarium head blight using UAV-based spectral and image feature fusion

Contact Info

Product

Resources

About