China
Africa
Explore chapters and articles related to this topic
Computing using Python Modules
Published in Ravishankar Chityala, Sridevi Pudipeddi, Image Processing and Acquisition using Python, 2020
Ravishankar Chityala, Sridevi Pudipeddi
In this book, we will focus only on the image processing routines in scikits named scikit-image. The scikit-image routine contains algorithms for input/output, morphology, object detection and analysis, etc. >>> from skimage import filters >>> import skimage.filters as fi In the first command, only the filters submodule is loaded. In the second command, the filters module is loaded as fi.
Image processing-based classification of pavement fatigue severity using extremely randomized trees, deep neural network, and convolutional neural network
Published in International Journal of Pavement Engineering, 2023
Nhat-Duc Hoang, Van-Duc Tran, Xuan-Linh Tran
It is worth noticing that the features related to the GSF-based PIs and the statistical measurement of three color channels are constructed in Python with the Microsoft Visual Studio integrated development environment by the authors. The features extracted from the GLCM are computed with the assistance of built-in functions provided in the scikit-image library (Walt, 2014). The feature computation processes for the image samples in three classes of interest are demonstrated in Figures 10–12. Given an input sample, the Gaussian steerable filters are applied to the image to characterise the edge-related features with respect to the horizontal (β = 0o) and vertical (β = 90o) directions. The statistical descriptions of the captured edges are subsequently summarised by the use of HPI and VPI. In addition to the edge-related features, the surface condition of the pavement area with respect to textural information is described by the image’s color and the GLCM’s indices.
Determination of 3D pore network structure of freeze-dried maltodextrin
Published in Drying Technology, 2022
M. Thomik, S. Gruber, P. Foerst, E. Tsotsas, N. Vorhauer-Huget
The complete image processing procedure was performed with MATLAB (Math Works, USA, v. R2020a) using the image processing toolbox and Python 3.7.3, employing the libraries scikit-image,[23] SciPy,[24] and PoreSpy.[25] An overview of the image processing steps is provided in Figure 2. Various different possibilities for image filtering, segmentation, etc., are generally available in the conventional image processing tools. In the relevant literature related to the analysis of porous media several options are usually presented, e.g.[3,14,35,37,47,48] In this article, two very common methods, which are in detail Otsu thresholding and adaptive thresholding, are applied[44,45] in combination with common image filtering methods, which are Gaussian filter and anisotropic diffusion filter,[35,42,43] following the suggestions in Kaestner et al.[35] The comparison of the resulting data is used to evaluate the available options in regard of a suitable method for image analysis of the maltodextrin sample.
Remote sensing of mangroves using unmanned aerial vehicles: current state and future directions
Published in Journal of Spatial Science, 2021
Edward Zimudzi, Ian Sanders, Nicholas Rollings, Christian W. Omlin
These image processing software systems have provided various machine learning algorithms like SVM (Heumann 2011, Kumar and Patnaik 2013, Heenkenda et al. 2014), Random forest (Pham and Yoshino 2016), ANN (Wong and Fung 2014), maximum likelihood classifiers (Roslani et al. 2013, Giri et al. 2014, Wong and Fung 2014, Zhang et al. 2016) and decision trees (Heumann 2011, Wong and Fung 2014, Zhang et al. 2017). Of note is that most of the studies have used more than one technique to discriminate land covers on one image, and field studies are almost always carried out to support remote sensing. Other software, including open source Scikit-Learn, Scikit-Image, and R and Python programming have not been widely used mainly due to the time required to learn how to implement solutions within them, and in the case of MATLAB, the required costs and learning time. They have sophisticated routines that can be used for most of the image processing and machine learning functions, but require sophisticated knowledge, and cannot be used by researchers not well-versed in machine learning and image processing.