All Projects → Grootzz → Glcm Svm

Grootzz / Glcm Svm

提取图像的灰度共生矩阵(GLCM),根据GLCM求解图像的概率特征,利用特征训练SVM分类器,对目标分类

Programming Languages

matlab
3953 projects

Projects that are alternatives of or similar to Glcm Svm

Sinaweibo Emotion Classification
新浪微博情感分析应用
Stars: ✭ 118 (+145.83%)
Mutual labels:  classification, svm
Augmentedgaussianprocesses.jl
Gaussian Process package based on data augmentation, sparsity and natural gradients
Stars: ✭ 99 (+106.25%)
Mutual labels:  classification, svm
Gru Svm
[ICMLC 2018] A Neural Network Architecture Combining Gated Recurrent Unit (GRU) and Support Vector Machine (SVM) for Intrusion Detection
Stars: ✭ 76 (+58.33%)
Mutual labels:  classification, svm
Tensorflow cookbook
Code for Tensorflow Machine Learning Cookbook
Stars: ✭ 5,984 (+12366.67%)
Mutual labels:  classification, svm
dzetsaka
dzetsaka : classification plugin for Qgis
Stars: ✭ 61 (+27.08%)
Mutual labels:  svm, classification
Tiny ml
numpy 实现的 周志华《机器学习》书中的算法及其他一些传统机器学习算法
Stars: ✭ 129 (+168.75%)
Mutual labels:  classification, svm
Nlp Journey
Documents, papers and codes related to Natural Language Processing, including Topic Model, Word Embedding, Named Entity Recognition, Text Classificatin, Text Generation, Text Similarity, Machine Translation),etc. All codes are implemented intensorflow 2.0.
Stars: ✭ 1,290 (+2587.5%)
Mutual labels:  classification, svm
Ml Course
Starter code of Prof. Andrew Ng's machine learning MOOC in R statistical language
Stars: ✭ 154 (+220.83%)
Mutual labels:  classification, svm
Fuku Ml
Simple machine learning library / 簡單易用的機器學習套件
Stars: ✭ 280 (+483.33%)
Mutual labels:  classification, svm
Jsmlt
🏭 JavaScript Machine Learning Toolkit
Stars: ✭ 22 (-54.17%)
Mutual labels:  classification, svm
Constrained attention filter
(ECCV 2020) Tensorflow implementation of A Generic Visualization Approach for Convolutional Neural Networks
Stars: ✭ 36 (-25%)
Mutual labels:  classification
Ml Classification Algorithms Poverty
A comparative assessment of machine learning classification algorithms applied to poverty prediction
Stars: ✭ 36 (-25%)
Mutual labels:  classification
Small norb
Python wrapper to small NORB dataset
Stars: ✭ 40 (-16.67%)
Mutual labels:  classification
Yolo tensorflow
🚖 Object Detection (YOLOv1) implentation in tensorflow, with training, testing and video features.
Stars: ✭ 45 (-6.25%)
Mutual labels:  classification
Satellite imagery analysis
Implementation of different techniques to find insights from the satellite data using Python.
Stars: ✭ 31 (-35.42%)
Mutual labels:  classification
Chemometricstools.jl
A collection of tools for chemometrics and machine learning written in Julia.
Stars: ✭ 39 (-18.75%)
Mutual labels:  classification
The Deep Learning With Keras Workshop
An Interactive Approach to Understanding Deep Learning with Keras
Stars: ✭ 34 (-29.17%)
Mutual labels:  classification
Patternrecognition matlab
Feature reduction projections and classifier models are learned by training dataset and applied to classify testing dataset. A few approaches of feature reduction have been compared in this paper: principle component analysis (PCA), linear discriminant analysis (LDA) and their kernel methods (KPCA,KLDA). Correspondingly, a few approaches of classification algorithm are implemented: Support Vector Machine (SVM), Gaussian Quadratic Maximum Likelihood and K-nearest neighbors (KNN) and Gaussian Mixture Model(GMM).
Stars: ✭ 33 (-31.25%)
Mutual labels:  svm
Prediciting Binary Options
Predicting forex binary options using time series data and machine learning
Stars: ✭ 33 (-31.25%)
Mutual labels:  classification
Detext
DeText: A Deep Neural Text Understanding Framework for Ranking and Classification Tasks
Stars: ✭ 1,039 (+2064.58%)
Mutual labels:  classification

基于灰度共生矩阵的图形纹理检测及路面状况的SVM分类实现

[TOC]

图像的特征提取是图像的识别和分类、基于内容的图像检索、图像数据挖掘等研究内容的基础性工作,其中图像的纹理特征对描述图像内容具有重要意义,纹理特征提取已成为图像领域研究的一个重要方法。

本项目以道路状况分类为背景,基于SVM使用纹理特征参数完成对道路状况的分类。

灰度共生矩阵的特征参数

  • 二阶矩
  • 对比度
  • 相关性
  • 逆差距

实现过程

本文以识别路况为背景设计系统,首先读取图像文件,在为了得到较为理想的结果,给出的图片尺寸较小,这样的目的在于提高运行速度、能够对理想情况下的识别结果有更加精确的把握。由于RGB图像的体积较大,但另一方面我们也知道,一张图片也可以用其灰度图像来反映,因此,对目标图像进行灰度的转化,可以降低对设计复杂度和运行效率的要求。尽管灰度转化后可以有效的降低图片的大小,但是对于256灰度级的图片仍然会给处理带来一些时间上的消耗,因此需要在保持原图像信息不变的情况下大量削减灰度级,通过对图像进行直方图均衡化处理,增加灰度值的动态范围,从而增加图像的整体对比效果。通常,需要将每个扫描的方向所计算的特征参数计算出来作一个权衡,这样能够更好地综合影响因素,一般会将每个方向的特征参数计算出来取平均,这样就可以以一个综合的指标来识别图像了。把所有的设计思路综合起来,可以用下图所示的系统设计框图来反映。

1555226955414.png

分类结果

通过样本标签和分类后的标签对比,得出使用能量和对比度作为训练分类器的样本,分类器对于实验待分类样本分类的其正确率为:95%,由此可以看出,这个分类器对于道路状况的分类仍然具有一定的意义。

1555226757177.png

设计总结

  • 计算灰度共生矩阵时要考虑三个变量,图像灰度级、方向和距离。

  • 对于灰度级的选取,灰度级决定了灰度共生矩阵的计算规模,降低灰度级可以提高计算速度和减少存储空间需求,且适当降低灰度级还可以减少噪声对图像的影响,但过小的灰度级会破坏有用纹理的成分。

  • 对于距离d的选择,共生矩阵在精细纹理中随距离而快速变化,而在粗糙纹理中随距离则变化缓慢。一般而言,对于平滑纹理用较大的距离,对于粗糙纹理用较小的距离会取得较好的效果。对于方向的选择,一般有四种取值,通过不同方向可以考察不同的纹理,不同方向生成的共生矩阵中包含不同的纹理信息。在考虑方向时,往往是分别计算四个方向灰度共生矩阵所确定的纹理特征值,然后以各方向特征值的均值作为最终纹理分量。

  • 利用SVM进行分类时,样本的选取和核函数的选取是分类器是否优良的一个重要影响因素,因此,对于一个优良的分类器,应该充分考虑各种影响因素后,选择一个最优的方案才能使待分类目标得到更好的分类。

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].