Ingo's Blog

( German ) This post describes a Matlab function which load out of a folder all songs, process them to a standarized name and save them in a folder in the different interpret folders. Mode of operation: First of all the path is chosen where the songs are, which should be sorted and then the folder is chosen, where the songs are copied to. Because in many folders are files like "..", ".", "desktop.ini", these are filtered. So now are all names of the file loaded. Next the name (interpret and songname) is devided and also further informations in the name like "(Official Video)" or "[pleer.net]" are cutted out.  Now the first song is taken and if there is a folder existing in the folder where it should be saved, then the song is saved, if it isn't already in the folder. If the interpret folder is not existing right now a new folder is generated with the name of the interpret and the song is copied into it. This is done for al...

( German ) This post is about a Matlab function where one can load an images and remove the background, by saving it as a transparent image. The background can consist out of different and blurring colors! Useful for prensentations that the image fits to the slice background. Mode of operation: This is a GUI based function, so that one can load via interface an images of any type. Now one can use the buttom "Choose BG" to select the background color in an additional appearing image. If it is chosen one see in figure 2 the extracted regions in white and in the lower left corner which color was selected. If one uses now the slider one can increase the range of the color. If the value is 10 and the chosen background color is [50 5 200] then the algorithm is also selecting every pixel with the colors between [40 0 190] and [60 15 210]. This is also seen in the lower left corner.  The buttom "next BG" can be used to select a second additional backgro...

( German ) This post is like the post before and describes the identification of the numbers on a twelve-sided dice with numbers from 1 to 12. Also automatically rolled by the dice rolling machine of  my brother . Dice identification: First of all the images are loaded into Matlab and the size is reduces to 200x200 pixel size. The with rgb2gray the image is reduces to a gray-scaled image and the gradient is used. The gradient and the normal image is converted into a binare image with a treshold. Then the gradient images is morphologically closed and multiplied with the normal tresholded image and afterwards morphologically dilated. After this the big for loop over all images is  executed. The the watershed is applied and the middle of the image with my own function is determined, which just calculate the mean of all columns and rows where the image is one. From this middle a big area is chosen that the rolled number is for sure in this area. Then the bigges...

( German ) This post describes the segmentation of images of dice with plus, minus and nothing on it Dice identification: My brother has built a dice rolling machine which automatically roll dices with 2/6 +, 1/6 - and 3/6 nothing on it. The images he produces automatically looked like this:  Here one can see that the + is in green and the - in red. Image Processing: First of all I loaded all images into Matlab and have cut them to a 200x200 size (second column). Further on I made a better contrast by multiplying all rgb-values with a factor, because the green plus and red minus had just a value of around 100 of 255 in rgb (third column). Then I splitted the channels in red, green and blue and generated out of this 3 channels on channel only red and only grenn by taken the red channel and subtracting the blue channel and the green channel with the factor 0.5. With a treshold in these images of 50 for green and 20 for red respectively I converted the images to binar...

Segmentation Growing ( German ) This post is about a methode of segmentation where a region is defined (seed) . In this seed the standard deviation and the mean is calculated and with a factor f it is weigthed. Pixel which are in the neighborhood of the seed are added to the seed if their value is in the boundary defined by the standard deviation, mean value and weighting factor f. This is done iterative as long as no more pixel is added to the seed. Then the object is segmented. function [region, area, seed, sx1, sx2, sy1, sy2] = Segmentation_Growing(img,sx1,sx2,sy1,sy2,f): This function needs an image as an imput. If there is no other input value one can choose with to cursors the starting point sx1, sy1, the left upper corner, and the end point sx2, sy2, the right lower corner. Otherwise one can define the s-value as one wants. If one uses cursors, one can use the second input argument for defining the factor f. This factor describes the weigthing and is in an area bigger equa...

The Dijkstra Methode - large neighborhood ( German )This is an update for the Dijkstra methode from the post before, where it uses now a larger neighborhood. It needs a bit longer but the probability to find the shortest path of many possible paths in the neighborhood is much higher. Here the comparing picture to Dijkstra Methode . Here one can download the following code on MathWorks: Source Code Dijkstra: function [path, prev, unvis, distance, start, target] = Dijkstra_Methode(Matrix, start, target) %Matrix is the incoming image %start is the start point in a vector [a,b] where a is the column and b the %row %target is the end point similare to start %path is the matrix with ones excepted at the position of the path where it is 0 %prev are also the previous visited pixels where the algorithm took the %wrong way %unvis are all unvisited pixels %distance is the distance or weight of the pixels %open the image and you can choose with the cursor the start %and end point if you d...

The Dijkstra Methode ( German ) In this post i show you with Matlab how to calculate the shortest path on a gradient images with d´the dijkstra methode i wrote. I take an image as an input and convert it to an double and calculate the gradient [G,Gdir] = imgradient(img,'sobel'); and I use my function Dijkstra(G,p,q) : THis function needs a image as double G , a start value p and the target q , where the first entry of the oints is the column and the second the row. This methode sets all distances to infinity except the at the start value is zero obviously. All unvisited pixels have the value 1. The algortihm continous as long as the target is an unvisible pixel. For all neighborhood pixels the weights are calculated with the function [weight] = localedgewidth(G,p,q) ,where the value of the neightbor pixel is substract by the maximum of the matrix and devided by the difference of maximal and minimal value of the matrix G and multiplied with the distance to the neighbor pi...