Basakstat - Image Analysis and Classification with R

Image Analysis and Classification with R

Chapter 1 Image or picture analysis and processing with R- reading and writing picture file- intensity histogram- combining images- merging images into one picture- image manipulation (brightness, contrast, gamma correction, cropping, color change, flip, flop, rotate, & resize)- low-pass and high pass filter Chapter 2 Image Classification & Recognition with Keras - load keras and EBImage packages- read images- explore images and image data- resize and reshape images- one hot encoding- sequential model- compile model- fit model- evaluate model- prediction- confusion matrix Chapter 3 Convolutional Neural Network wirh Keras & TensorFlow in R - Advantages- layers- parameter calculations- load keras and EBImage packages- read images- explore images and image data- resize and reshape images- one hot encoding- sequential model- compile model- fit model- evaluate model- prediction- confusion matrix Chapter 4 Image Recognition Using Google Cloud Vision API in R Chapter 5 Live Face Detection

Chapter 1: Image or picture analysis and processing with R

Code:# Install EBImagesource("http://bioconductor.org/biocLite.R")biocLite()biocLite("EBImage") # Read Imagelibrary(EBImage)Image1 <- readImage("~/Desktop/pic1.jpg")Image2 <- readImage("~/Desktop/pic2.jpg") # Plot datahist(Image1) # Manipulating brightness# Lighta <- Image1 + 0.4print(a)display(a) # Darkb <- Image1 - 0.4display(b)hist(b) # Combinec <- combine(Image1,Image2)display(c) # Put two picture into oned <- Image1 + Image2/4display(d)hist(d) # Manipulating contraste <- Image1*0.5display(e)f <- Image1*3display(f) # Gamma Correctiong <- Image1^0.5h <- Image1^3display(g)display(h) # ColorcolorMode(Image1) <- Grayscaleprint(Image1)display(Image1) colorMode(Image1) <- Color #to return to color # Croppingk <- Image1[800:2686, 800:2200,]display(k) # New image filewriteImage(k, "NewImage.jpg") # Flip, Flop, Rotate, resizel <- flip(Image1)display(l)m <- rotate(Image1, 45)display(m)n<- flop(Image1)display(n)o <- resize(Image1, 400)display(o) # Low-pass filterlow <- makeBrush(81, shape = 'disc', step=FALSE)^2low <- low/sum(low)Image.low <- filter2(Image1, low)display(Image.low ) # High-pass filterhigh <- matrix(1, nc=3, nr = 3)high[2,2] <- -5Image.high <- filter2(Image1, high)display(Image.high) # Combinenew<- Image.high/5+Image1comb <- combine(Image1, new)display(comb)hist(comb)

Chapter 2: Image Classification & Recognition with Keras

Code:# Load Packageslibrary(EBImage)library(keras) # To install EBimage package, you can run following 2 lines;# install.packages("BiocManager")# BiocManager::install("EBImage") # Read imagessetwd('---')pics <- c('p1.jpg', 'p2.jpg', 'p3.jpg', 'p4.jpg', 'p5.jpg', 'p6.jpg', 'c1.jpg', 'c2.jpg', 'c3.jpg', 'c4.jpg', 'c5.jpg', 'c6.jpg')mypic <- list()for (i in 1:12) {mypic[[i]] <- readImage(pics[i])} # Exploreprint(mypic[[1]])display(mypic[[8]])summary(mypic[[1]])hist(mypic[[2]])str(mypic) # Resizefor (i in 1:12) {mypic[[i]] <- resize(mypic[[i]], ---, ---)} # Reshapefor (i in 1:12) {mypic[[i]] <- array_reshape(mypic[[i]], c(---, ---,---))} # Row Bindtrainx <- NULLfor (i in 7:11) {trainx <- rbind(trainx, mypic[[i]])}str(trainx)testx <- ---(mypic[[6]], mypic[[12]])trainy <- c(0,0,0,0,0,1,1,1,1,1 )testy <- c(---, ---) # One Hot EncodingtrainLabels <- ---(trainy)testLabels <- ---(testy) # Modelmodel <- keras_model_sequential()model %>% layer_dense(units = 256, activation = ---, input_shape = c(2352)) %>% layer_dense(units = 128, activation = 'relu') %>% layer_dense(units = 2, activation = ---)summary(model) # Compilemodel %>% compile(loss = ---, optimizer = optimizer_rmsprop(), metrics = c('accuracy')) # Fit Modelhistory <- model %>% fit(trainx, ---, epochs = 30, batch_size = 32, validation_split = 0.2) # Evaluation & Prediction - train datamodel %>% evaluate(---, ---)pred <- model %>% predict_classes(trainx)table(Predicted = pred, Actual = trainy)prob <- model %>% predict_proba(trainx)cbind(prob, Prected = pred, Actual= trainy)

Chapter 3: Convolutional Neural Network wirh Keras & TensorFlow in R

Advantages:1. Regarded as ‘Gold Standard’ for image classification2. Less parameters needed compared to a fully connected network.3. Therefore, less processing time4. Suitable for image recognition and classification with large number of images. Layers:1. Input Layer2. Convolutional Layer3. Pooling Layer4. Dropout Layer5. Flattening Layer6. Fully connected Layer7. Output Layer Code:# Load packageslibrary(keras)library(EBImage) # Read Imagessetwd('---')pic1 <- c('p1.jpg', 'p2.jpg', 'p3.jpg', 'p4.jpg', 'p5.jpg', 'c1.jpg', 'c2.jpg', 'c3.jpg', 'c4.jpg', 'c5.jpg', 'b1.jpg', 'b2.jpg', 'b3.jpg', 'b4.jpg', 'b5.jpg')train <- list()for (i in 1:15) {train[[i]] <- readImage(pic1[i])} pic2 <- c('p6.jpg', 'c6.jpg', 'b6.jpg')test <- list()for (i in 1:3) {test[[i]] <- readImage(pic2[i])} # Exploreprint(train[[12]])summary(train[[12]])display(train[[12]])plot(train[[12]]) par(mfrow = c(---, ---))for (i in 1:15) plot(train[[i]])par(mfrow = c(1,1)) # Resize & combinestr(train)for (i in 1:15) {train[[i]] <- resize(train[[i]], 100, 100)}for (i in 1:3) {test[[i]] <- ---(test[[i]], 100, 100)} train <- combine(train)x <- tile(train, 5)display(x, title='Pictures') test <- ---(test)y <- ---(test, 3)---(y, title = 'Pics') # Reorder dimensiontrain <- ---(train, c(4, 1, 2, 3))test <- ---(test, c(4, 1, 2, 3))str(train) # Responsetrainy <- c(0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2)testy <- c(0, 1, 2) # One hot encodingtrainLabels <- ---(trainy)testLabels <- ---(testy) # Modelmodel <- keras_model_sequential() model %>% layer_conv_2d(filters = 32, kernel_size = c(3,3), activation = '---', input_shape = c(---, ---, ---)) %>% ---(filters = 32, kernel_size = c(3,3), activation = 'relu') %>% ---(pool_size = c(2,2)) %>% ---(rate = 0.25) %>% layer_conv_2d(filters = 64, kernel_size = c(3,3), activation = 'relu') %>% layer_conv_2d(filters = 64, kernel_size = c(3,3), activation = 'relu') %>% layer_max_pooling_2d(pool_size = c(2,2)) %>% layer_dropout(rate = 0.25) %>% ---() %>% layer_dense(units = 256, activation = 'relu') %>% layer_dropout(rate=0.25) %>% layer_dense(units = 3, activation = '---') %>% compile(loss = '---', optimizer = optimizer_sgd(lr = 0.01, decay = 1e-6, momentum = 0.9, nesterov = T), metrics = c('---'))summary(model) # Fit modelhistory <- model %>% fit(train, trainLabels, --- = 60, --- = 32, validation_split = 0.2, validation_data = list(test, testLabels))plot(history) # Evaluation & Prediction - train datamodel %>% ---(train, trainLabels)pred <- model %>% ---(train)table(Predicted = pred, Actual = trainy) prob <- model %>% ---(train)cbind(prob, Predicted_class = pred, Actual = trainy) # Evaluation & Prediction - test datamodel %>% ---(test, testLabels)pred <- model %>% ---(test)table(Predicted = pred, Actual = testy) prob <- model %>% ---(test)---(prob, Predicted_class = pred, Actual = testy)

Chapter 4: Image Recognition Using Google Cloud Vision API in R

Code:# Librarieslibrary(tidyverse)library(leaflet)# devtools::install_github("flovv/RoogleVision)library(RoogleVision)library(jsonlite)# source("http://bioconductor.org/biocLite.R")# biocLite("EBImage")library(EBImage) # Google Cloudc <- fromJSON(file.choose())options("googleAuthR.client_id" = c$installed$client_id)options("googleAuthR.client_secret" = c$installed$client_secret)options("googleAuthR.scopes.selected" = c("https://www.googleapis.com/auth/cloud-platform"))googleAuthR::gar_auth() # Lable Detectionpic <- readImage(file.choose())plot(pic)p <- getGoogleVisionResponse(file.choose(), feature = "LABEL_DETECTION") # Landmark Detectionpic <- readImage(file.choose())plot(pic)p <- getGoogleVisionResponse(file.choose(), feature = "LANDMARK_DETECTION")for(i in 1:length(p$boundingPoly$vertices)){ a = p$boundingPoly$vertices[[i]]$x b = p$boundingPoly$vertices[[i]]$y polygon(x=a, y=b, border = "green", lwd=20)} # Mapla <- p$locations[[1]][[1]][[1]]lo <- p$locations[[1]][[1]][[2]]map <- leaflet() %>% addProviderTiles(providers$CartoDB.Positron) %>% setView(lng = lo, lat = la, zoom = 30) %>% addMarkers(lng = lo, lat = la)map # Face Detectionpic <- readImage(file.choose())plot(pic)p <- getGoogleVisionResponse(file.choose(), feature = "FACE_DETECTION")for(i in 1:length(p$boundingPoly$vertices)){ a = p$boundingPoly$vertices[[i]]$x b = p$boundingPoly$vertices[[i]]$y polygon(x=a, y=b, border = "green", lwd=5)}a <- p$landmarks[[1]][[2]][[1]]b <- p$landmarks[[1]][[2]][[2]]points(x=a, y=b, lwd = 2, col = "red") # Logo Detectionpic <- readImage(file.choose())plot(pic)p <- getGoogleVisionResponse(file.choose(), feature = "LOGO_DETECTION")a <- p$boundingPoly$vertices[[1]][[1]]b <- p$boundingPoly$vertices[[1]][[2]]polygon(x=a, y=b, border="green", lwd=20) # Text Detectionpic <- readImage(file.choose())plot(pic)p <- getGoogleVisionResponse(file.choose(), feature = "TEXT_DETECTION")for(i in 1:length(p$boundingPoly$vertices)){ a = p$boundingPoly$vertices[[i]]$x b = p$boundingPoly$vertices[[i]]$y polygon(x=a, y=b, border = "red", lwd=5)}

Chapter 5: Live Face Detection

Code:#Librarieslibrary(opencv)library(psych) # Live face detection:ocv_video(ocv_face) # Various optionsocv_video(ocv_edges)ocv_video(ocv_knn)ocv_video(ocv_facemask)ocv_video(ocv_mog2)ocv_video(ocv_stylize)ocv_video(ocv_sketch) # Overlay face filter test <- ocv_camera() bitmap <- ocv_bitmap(test)width <- dim(bitmap)[2]height <- dim(bitmap)[3] png('bg.png', width = width, height = height)data('iris')print(pairs.panels(iris[1:4], gap=0, pch=21, bg = c("red", "green", "blue")[iris$Species]))dev.off() bg <- ocv_read('bg.png') ocv_video(function(input){ mask <- ocv_facemask(input) ocv_copyto(input, bg, mask) }) # Face recognitionccb <- ocv_read('~/Desktop/ccb.jpg')faces <- ocv_face(ccb)ocv_write(b, '~/Desktop/b.jpg') # Various optionsocv_sketch(ccb, color = T)ocv_blur(ccb, ksize = 15)ocv_hog(ccb)ocv_markers(ccb)ocv_stylize(ccb) # get the face location data:facemask <- ocv_facemask(ccb)attr(facemask, 'faces')

Page updated

Google Sites

Report abuse