In Selenium, how to compare images? - image

As i have been asked to automate our Company's website using Selenium Automation tooL.
But i am new to Selenium tool to proceed with, but i have learnt the basics of Selenium IDE and RC. But i am very much confused with how to compare actual and original images as we usually do in other automation tools. How do we come to a result that there bug in the website? Its obviously through image comparison but i wonder as selenium is one of the very popular tools but it doesn't have image comparing option. On the other hand i doubt whether my way of proceeding with the automation process is correct! Could somebody please help me out..
Thanks in Advance!!
Sanjay S

I had simillar task. I needed to compare more than 3000 images on a WebPage.
First of all I scrolled page to load all images:
public void compareImage() throws InterruptedException {
driver.get(baseUrl);
driver.manage().window().maximize();
JavascriptExecutor executor = (JavascriptExecutor) driver;
Long previousHeight;
Long currentHeight;
do {
previousHeight = (Long) executor.executeScript("return document.documentElement.scrollHeight");
executor.executeScript("window.scrollBy(0, document.documentElement.scrollHeight)");
Thread.sleep(500);
currentHeight = (Long) executor.executeScript("return document.documentElement.scrollHeight");
} while (Long.compare(previousHeight, currentHeight) != 0);
after I compared size of all images with first image(or you can just write size):
List<WebElement> images = driver.findElements(By.cssSelector("img[class='playable']"));
List<String> errors = new LinkedList<>();
int imgWidth, imgHeight, elWidth, elHeight;
int imgNum = 0;
imgWidth = images.get(0).getSize().getWidth();
imgHeight = images.get(0).getSize().getHeight();
for (WebElement el : images) {
imgNum++;
elWidth = el.getSize().getWidth();
elHeight = el.getSize().getHeight();
if (imgWidth != elWidth || imgHeight != elHeight) {
errors.add(String.format("Picture # %d has incorrect size (%d : %d) px"
, imgNum, elWidth, elHeight));
}
}
for (String str : errors)
System.out.println(str);
if (errors.size() == 0)
System.out.println("All images have the same size");
}

Since you mention knowledge about Selenium RC, you can easily extend Selenium's capability using a library for your chosen programming language. For instance, in Java you can use the PixelGrabber class for comparing two images and assert their match.

imagemagick and imagediff are also two good tools to use for image matching. You would require Selenium RC and a programming language knowledge to work with it.

Image comparison on C#. To get exact results I recommend to disable anti aliasing browser feature before taking screenshots, otherwise pixels each time are a little bit different drawn. For example HTML canvas element options.AddArgument("disable-canvas-aa");
private static bool ImageCompare(Bitmap bmp1, Bitmap bmp2, Double TolerasnceInPercent)
{
bool equals = true;
bool flag = true; //Inner loop isn't broken
//Test to see if we have the same size of image
if (bmp1.Size == bmp2.Size)
{
for (int x = 0; x < bmp1.Width; ++x)
{
for (int y = 0; y < bmp1.Height; ++y)
{
Color Bitmap1 = bmp1.GetPixel(x, y);
Color Bitmap2 = bmp2.GetPixel(x, y);
if (Bitmap1.A != Bitmap2.A)
{
if (!CalculateTolerance(Bitmap1.A, Bitmap2.A, TolerasnceInPercent))
{
flag = false;
equals = false;
break;
}
}
if (Bitmap1.R != Bitmap2.R)
{
if (!CalculateTolerance(Bitmap1.R, Bitmap2.R, TolerasnceInPercent))
{
flag = false;
equals = false;
break;
}
}
if (Bitmap1.G != Bitmap2.G)
{
if (!CalculateTolerance(Bitmap1.G, Bitmap2.G, TolerasnceInPercent))
{
flag = false;
equals = false;
break;
}
}
if (Bitmap1.B != Bitmap2.B)
{
if (!CalculateTolerance(Bitmap1.B, Bitmap2.B, TolerasnceInPercent))
{
flag = false;
equals = false;
break;
}
}
}
if (!flag)
{
break;
}
}
}
else
{
equals = false;
}
return equals;
}

This C# function calculates tolerance
private static bool CalculateTolerance(Byte FirstImagePixel, Byte SecondImagePixel, Double TolerasnceInPercent)
{
double OneHundredPercent;
double DifferencesInPix;
double DifferencesPercentage;
if (FirstImagePixel > SecondImagePixel)
{
OneHundredPercent = FirstImagePixel;
}
else
{
OneHundredPercent = SecondImagePixel;
}
if (FirstImagePixel > SecondImagePixel)
{
DifferencesInPix = FirstImagePixel - SecondImagePixel;
}
else
{
DifferencesInPix = SecondImagePixel - FirstImagePixel;
}
DifferencesPercentage = (DifferencesInPix * 100) / OneHundredPercent;
DifferencesPercentage = Math.Round(DifferencesPercentage, 2);
if (DifferencesPercentage > TolerasnceInPercent)
{
return false;
}
return true;
}

Related

How do the obstacle detection sensors work in the DJI Windows SDK?

I want to get the distance of the obstacles in the back and front of the Mavic Air.
I'm using the VissionDetectionStateChanged and it returns values in the 4 sectors, but all of them change only with obstacles in the back. If I put my hand in the front, nothing happens.
The VisionSensorPosition is returning TAIL always and when I put my hand very close to the tail of the aircraft it changes for NOSE.
Shouldn't be the opposite?
Right now I just display the information, but I'd like to be able to detect obstacles in the back and front of the aircraft to try to keep it in the middle of two objects and avoid collisions.
This is my code in the event:
private async void FlightAssistant_VissionDetectionStateChanged(object sender, VissionDetectionState? value)
{
if (value.HasValue)
{
if (txtPosition.Dispatcher.HasThreadAccess)
{
txtPosition.Text = value.Value.position.ToString();
for (int i = 0, count = value.Value.detectionSectors.Count; i < count; i++)
{
ObstacleDetectionSector sector = value.Value.detectionSectors[i];
TextBox txtWarning = this.FindControl<TextBox>("txtWarning" + i.ToString());
if (txtWarning != null)
txtWarning.Text = sector.warningLevel.ToString();
TextBox txtObstacleDistance = this.FindControl<TextBox>("txtObstacleDistance" + i.ToString());
if (txtObstacleDistance != null)
txtObstacleDistance.Text = sector.obstacleDistanceInMeters.ToString();
}
}
else
{
await txtPosition.Dispatcher.RunAsync(Windows.UI.Core.CoreDispatcherPriority.Normal, () =>
{
txtIsSensorBeingUsed.Text = value.Value.isSensorBeingUsed.ToString();
txtPosition.Text = value.Value.position.ToString();
for (int i = 0, count = value.Value.detectionSectors.Count; i < count; i++)
{
ObstacleDetectionSector sector = value.Value.detectionSectors[i];
TextBox txtWarning = this.FindControl<TextBox>("txtWarning" + i.ToString());
if (txtWarning != null)
txtWarning.Text = sector.warningLevel.ToString();
TextBox txtObstacleDistance = this.FindControl<TextBox>("txtObstacleDistance" + i.ToString());
if (txtObstacleDistance != null)
txtObstacleDistance.Text = sector.obstacleDistanceInMeters.ToString();
}
});
}
}
}

QLineEdit and QComboBox have unrelated behavior

I have a user interface with a lot of controls. However I have a problem with a QLineEdit and a QComboBox that are not responding properly.
I am basically converting from pixel measurements to millimeters/centimeters/decimeters and meters with a QComboBox and showing the result on a QLineEdit.
For the conversion table I used this page.
When I choose fromPixelToMillimeters() it does the conversion, but when I choose fromPixelToCentimeters() I think it is using the present value after the first conversion of fromPixelToMillimeters(). And if I go back choosing fromPixelToMillimeters() I get a different result too. This happens continuously, I get different measures each time.
See the code below:
void MainWindow::on_cBoxMeasures_currentIndexChanged(const QString &arg1)
{
if(arg1 == "Select Conversion(s)") {
return ui->leftLineEditDist->setText(QString("%1").arg(ui->leftLineEditDist->text().toDouble()));
} else if(arg1 == "pixel") {
return ui->leftLineEditDist->setText(QString("%1").arg(ui->leftLineEditDist->text().toDouble()));
} else if(arg1 == "mm") {
return fromPixelToMillimeters();
} else if(arg1 == "dm") {
return fromPixelToDecimeters();
} else if(arg1 == "cm") {
return fromPixelToCentimeters();
} else if(arg1 == "m") {
return fromPixelToMeters();
}
}
void MainWindow::fromPixelToMillimeters()
{
double mm = ui->leftLineEditDist->text().toDouble();
double dpi = 300;
double totalDistanceInMillimeter = (mm*25.4)/dpi;
ui->leftLineEditDist->setText(QString("%1").arg(totalDistanceInMillimeter));
ui->leftLineEditDist->show();
}
void MainWindow::fromPixelToCentimeters()
{
double mm = ui->leftLineEditDist->text().toDouble();
double dpi = 300;
double totalDistanceInCm = ((mm*25.4)/dpi)*0.1;
ui->leftLineEditDist->setText(QString("%1").arg(totalDistanceInCm));
ui->leftLineEditDist->show();
}
void MainWindow::fromPixelToDecimeters()
{
double mm = ui->leftLineEditDist->text().toDouble();
double dpi = 300;
double totalDistanceInDcm = ((mm*25.4)/dpi)*0.01;
ui->leftLineEditDist->setText(QString("%1").arg(totalDistanceInDcm));
ui->leftLineEditDist->show();
}
void MainWindow::fromPixelToMeters()
{
double mm = ui->leftLineEditDist->text().toDouble();
double dpi = 300;
double totalDistanceInM = ((mm*25.4)/dpi)*0.001;
ui->leftLineEditDist->setText(QString("%1").arg(totalDistanceInM));
ui->leftLineEditDist->show();
}
void MainWindow::on_cBoxMeasures_currentIndexChanged(int index)
{
switch (index) {
case(0):
break;
case(1):
break;
case(2):
fromPixelToMillimeters();
break;
case(3):
fromPixelToCentimeters();
break;
case(4):
fromPixelToDecimeters();
break;
case(5):
fromPixelToMeters();
break;
}
}
Please advise on what the problem might be.
I think these slots
on_cBoxMeasures_currentIndexChanged(const QString &arg1)
on_cBoxMeasures_currentIndexChanged(int index)
are connected the onIndexChange signal.
When the combo value will be changed, these two slots will be called simultaneously.
So that your code wont work well.
I recommend you to remove one of these slots.

Path finding with theta* when the triangle inequality is not fulfilled

I try to use the theta* algorithm (aigamedev.com/open/tutorial/lazy-theta-star) to find the fastest path on a rectangular grid. Distances are euqlidian, but speed between nodes is time dependent and varies between directions. So the triangle inequality, which is a prerequisite for the algorithm, is violated. Still it works excellently in most cases. How could I modify the code to work nicely also in turbulent areas? I suspect I may have to reevaluate some closed nodes and put them back into the open list. If so, under what conditions? An extensive web search hasn't helped.
vertex_t *thetastar(vertex_t *startnode, vertex_t *finishnode, double starttime) {
vertex_t *s, *s1;
double gold, gnew;
int dir; //8 directions to search for node neighbours
//Initialize
vertex[0].row = startnode->row;
vertex[0].col = startnode->col;
vertex[0].g = starttime;
vertex[0].h = h(&vertex[0], finishnode);
vertex[0].open = true;
vertex[0].closed = false;
vertex[0].parent = &vertex[0];
openlist[0] = &vertex[0];
openlist[1] = NULL;
//Find path
while ((s = pop(openlist)) != NULL) {
if (s->row == finishnode->row && s->col == finishnode->col) {
return s;
}
s->closed = true;
for (dir = 0; dir < 8; dir++) {
if ((s1 = nghbrvis(s, dir)) != NULL) {
if (!s1->closed) {
if (!s1->open) {
s1->g = inftime;
s1->parent = NULL;
}
gold = s1->g;
//Path 2
if (lineofsight(s->parent, s1)) {
gnew = (s->parent)->g + c(s->parent, s1);
if (gnew < s1->g) {
s1->parent = s->parent;
s1->g = gnew;
} }
//Path 1
gnew = s->g + c(s, s1);
if (gnew < s1->g) {
s1->parent = s;
s1->g = gnew;
}
if (s1->g < gold) {
s1->h = h(s1, finishnode);
if (s1->open)
remove(s1, openlist);
insert(s1, openlist);
} } } } }
return NULL;
}

How to correct loop counters for maze algorithm?

I have figured out how to move my character around the maze using the algorithm I have written, but the count is not figuring correctly. At the end of each row my character moves up and down several times until the count reaches the specified number to exit the loop, then the character moves along the next row down until it reaches the other side and repeats the moving up and down until the count reaches the specified number again. Can anyone help me find why my count keeps getting off? The algorithm and the maze class I am calling from is listed below.
public class P4 {
public static void main(String[] args) {
// Create maze
String fileName = args[3];
Maze maze = new Maze(fileName);
System.out.println("Maze name: " + fileName);
// Get dimensions
int mazeWidth = maze.getWidth();
int mazeHeight = maze.getHeight();
// Print maze size
System.out.println("Maze width: " + mazeWidth);
System.out.println("Maze height: " + mazeHeight);
int r = 0;
int c = 0;
// Move commands
while (true){
for (c = 0; c <= mazeWidth; c++){
if (maze.moveRight()){
maze.isDone();
c++;
}
if (maze.isDone() == true){
System.exit(1);
}
if (maze.moveRight() == false && c != mazeWidth){
maze.moveDown();
maze.moveRight();
maze.moveRight();
maze.moveUp();
c++;
}
}
for (r = 0; r % 2 == 0; r++){
maze.moveDown();
maze.isDone();
if (maze.isDone() == true){
System.exit(1);
}
}
for (c = mazeWidth; c >= 0; c--){
if (maze.moveLeft()){
c--;
maze.isDone();
System.out.println(c);
}
if (maze.isDone() == true){
System.exit(1);
}
if (maze.moveLeft() == false && c != 0){
maze.moveDown();
maze.moveLeft();
maze.moveLeft();
maze.moveUp();
c--;
}
}
for (r = 1; r % 2 != 0; r++){
maze.moveDown();
maze.isDone();
if (maze.isDone() == true){
System.exit(1);
}
}
}
}
}
public class Maze {
// Maze variables
private char mazeData[][];
private int mazeHeight, mazeWidth;
private int finalRow, finalCol;
int currRow;
private int currCol;
private int prevRow = -1;
private int prevCol = -1;
// User interface
private JFrame frame;
private JPanel panel;
private Image java, student, success, donotpass;
private ArrayList<JButton> buttons;
// Maze constructor
public Maze(String fileName) {
// Read maze
readMaze(fileName);
// Graphics setup
setupGraphics();
}
// Get height
public int getHeight() {
return mazeHeight;
}
// Get width
public int getWidth() {
return mazeWidth;
}
// Move right
public boolean moveRight() {
// Legal move?
if (currCol + 1 < mazeWidth) {
// Do not pass?
if (mazeData[currRow][currCol + 1] != 'D')
{
currCol++;
redraw(true);
return true;
}
}
return false;
}
// Move left
public boolean moveLeft() {
// Legal move?
if (currCol - 1 >= 0) {
// Do not pass?
if (mazeData[currRow][currCol - 1] != 'D')
{
currCol--;
redraw(true);
return true;
}
}
return false;
}
// Move up
public boolean moveUp() {
// Legal move?
if (currRow - 1 >= 0) {
// Do not pass?
if (mazeData[currRow - 1][currCol] != 'D')
{
currRow--;
redraw(true);
return true;
}
}
return false;
}
// Move down
public boolean moveDown() {
// Legal move?
if (currRow + 1 < mazeHeight) {
// Do not pass?
if (mazeData[currRow + 1][currCol] != 'D')
{
currRow++;
redraw(true);
return true;
}
}
return false;
}
public boolean isDone() {
// Maze solved?
if ((currRow == finalRow) && (currCol == finalCol))
return true;
else
return false;
}
private void redraw(boolean print) {
// Wait for awhile
try {
Thread.sleep(500);
} catch (InterruptedException ex) {
Thread.currentThread().interrupt();
}
if (print)
System.out.println("Moved to row " + currRow + ", column " + currCol);
// Compute index and remove icon
int index = (prevRow * mazeWidth) + prevCol;
if ((prevRow >= 0) && (prevCol >= 0)) {
buttons.get(index).setIcon(null);
}
// Compute index and add icon
index = (currRow * mazeWidth) + currCol;
if ((currRow == finalRow) && (currCol == finalCol))
buttons.get(index).setIcon(new ImageIcon(success));
else
buttons.get(index).setIcon(new ImageIcon(student));
// Store previous location
prevRow = currRow;
prevCol = currCol;
}
// Set button
private void setButton(JButton button, int row, int col) {
if (mazeData[row][col] == 'S') {
button.setIcon(new ImageIcon(student));
currRow = row;
currCol = col;
} else if (mazeData[row][col] == 'J') {
button.setIcon(new ImageIcon(java));
finalRow = row;
finalCol = col;
} else if (mazeData[row][col] == 'D') {
button.setIcon(new ImageIcon(donotpass));
}
}
// Read maze
private void readMaze(String filename) {
try {
// Open file
Scanner scan = new Scanner(new File(filename));
// Read numbers
mazeHeight = scan.nextInt();
mazeWidth = scan.nextInt();
// Allocate maze
mazeData = new char[mazeHeight][mazeWidth];
// Read maze
for (int row = 0; row < mazeHeight; row++) {
// Read line
String line = scan.next();
for (int col = 0; col < mazeWidth; col++) {
mazeData[row][col] = line.charAt(col);
}
}
// Close file
scan.close();
} catch (IOException e) {
System.out.println("Cannot read maze: " + filename);
System.exit(0);
}
}
// Setup graphics
private void setupGraphics() {
// Create grid
frame = new JFrame();
panel = new JPanel();
panel.setLayout(new GridLayout(mazeHeight, mazeWidth, 0, 0));
frame.add(Box.createRigidArea(new Dimension(0, 5)), BorderLayout.NORTH);
frame.add(panel, BorderLayout.CENTER);
// Look and feel
try {
UIManager.setLookAndFeel(UIManager.getCrossPlatformLookAndFeelClassName());
} catch (Exception e) {
e.printStackTrace();
}
// Configure window
frame.setSize(mazeWidth * 100, mazeHeight * 100);
frame.setTitle("Maze");
frame.setResizable(false);
frame.setLocationRelativeTo(null);
frame.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
frame.setAlwaysOnTop(true);
// Load and scale images
ImageIcon icon0 = new ImageIcon("Java.jpg");
Image image0 = icon0.getImage();
java = image0.getScaledInstance(100, 100, Image.SCALE_DEFAULT);
ImageIcon icon1 = new ImageIcon("Student.jpg");
Image image1 = icon1.getImage();
student = image1.getScaledInstance(100, 100, Image.SCALE_DEFAULT);
ImageIcon icon2 = new ImageIcon("Success.jpg");
Image image2 = icon2.getImage();
success = image2.getScaledInstance(100, 100, Image.SCALE_DEFAULT);
ImageIcon icon3 = new ImageIcon("DoNotPass.jpg");
Image image3 = icon3.getImage();
donotpass = image3.getScaledInstance(100, 100, Image.SCALE_DEFAULT);
// Build panel of buttons
buttons = new ArrayList<JButton>();
for (int row = 0; row < mazeHeight; row++) {
for (int col = 0; col < mazeWidth; col++) {
// Initialize and add button
JButton button = new JButton();
Border border = new LineBorder(Color.darkGray, 4);
button.setOpaque(true);
button.setBackground(Color.gray);
button.setBorder(border);
setButton(button, row, col);
panel.add(button);
buttons.add(button);
}
}
// Show window
redraw(false);
frame.setVisible(true);
}
}
One error I can see in your code is that you're incrementing your c counter more often than you should. You start with it managed by your for loop, which means that it will be incremented (or decremented, for the leftward moving version) at the end of each pass through the loop. However, you also increment it an additional time in two of your if statements. That means that c might increase by two or three on a single pass through the loop, which is probably not what you intend.
Furthermore, the count doesn't necessarily have anything obvious to do with the number of moves you make. The loop code will always increase it by one, even if you're repeatedly trying to move through an impassible wall.
I don't really understand what your algorithm is supposed to be, so I don't have any detailed advice for how to fix your code.
One suggestion I have though is that you probably don't ever want to be calling methods on your Maze class without paying attention to their return values. You have a bunch of places where you call isDone but ignore the return value, which doesn't make any sense. Similarly, you should always be checking the return values from your moveX calls, to see if the move was successful or not. Otherwise you may just blunder around a bunch, without your code having any clue where you are in the maze.

Can I get timecode from directshow video?

I'm trying to make a timecode counter for a video player based on GMFBridge and DirectShow.
I'm using a Timer to call GetCurrentPosition() every 200ms but I believe it's not accurate. I'd like at least to get the frame number (from start) of the current frame when a video is running.
Can this actually be done?
I'm using DirectShowLib .NET library.
To my knowledge this is hard to achieve, in a solution I work on I did the following to get 'frame number':
public int NumberOfFrames
{
get
{
return (int)(Duration / AverageTimePerFrame);
}
}
public double AverageTimePerFrame
{
get
{
return videoInfoHeader.AvgTimePerFrame / 10000000.0;
}
}
public int GetCurrentFrame(double currentTime)
{
int noOfFrames = (int)(Duration / AverageTimePerFrame);
return Convert.ToInt32(Math.Min(noOfFrames - 1, Math.Floor(currentTime / AverageTimePerFrame)));
}
I got the videoInfoHeader by doing:
// Get the media type from the SampleGrabber
AMMediaType media = new AMMediaType();
hr = sampGrabber.GetConnectedMediaType(media);
DsError.ThrowExceptionForHR(hr);
if ((media.formatType != FormatType.VideoInfo) || (media.formatPtr == IntPtr.Zero))
{
throw new NotSupportedException("Unknown Grabber Media Format");
}
// Grab the size info
videoInfoHeader = (VideoInfoHeader)Marshal.PtrToStructure(media.formatPtr, typeof(VideoInfoHeader));
DsUtils.FreeAMMediaType(media);
However this is obviously tailored to my own use-case, hopefully it helps you a bit though. Good luck!
Updated
Added CurrentTime code (the locker is for my own usage you can most likely remove that):
public double CurrentTime
{
set
{
lock (locker)
{
IMediaPosition mediaPos = fFilterGraph as IMediaPosition;
int hr;
if (value >= 0 && value <= Duration)
{
hr = mediaPos.put_CurrentPosition(value);
DsError.ThrowExceptionForHR(hr);
}
}
}
get
{
lock (locker)
{
IMediaPosition mediaPos = fFilterGraph as IMediaPosition;
int hr;
double currentTime;
hr = mediaPos.get_CurrentPosition(out currentTime);
DsError.ThrowExceptionForHR(hr);
return currentTime;
}
}
}

Resources