لخّصلي

?In past the traditional method to detecting brain tumours was by naked eye depends on area's and shape's of the tumour, here being the problem Cuz not all the doctors with same Knowledge, and the Eye diagnosis takes alot of time with many error and it decreased accurarcy of detection and Classification .After that use segmentation techniges to extract the features of ROI by used sobel oprator, simple it detecting edges by highlight areas of rapid intensity transitions.We used to parts in methodology the first one deals with detection the tumor , the other one Contains the Classification of tumor by matlab tools.So we had to Create system to detectd and classificate the tumor , to avoid using eye dignosis, reduce the time and Dignosis as many image as possible .

‏In past the traditional method to detecting brain tumours was by naked eye depends on area's and shape's of the tumour, here being the problem Cuz not all the doctors with same Knowledge, and the Eye diagnosis takes alot of time with many error and it decreased accurarcy of detection and Classification .
So we had to Create system to detectd and classificate the tumor , to avoid using eye dignosis, reduce the time and Dignosis as many image as possible .
We used to parts in methodology the first one deals with detection the tumor , the other one Contains the Classification of tumor by matlab tools.
First we Collect dataset from source then takes all image and filterate it with :
Median : minimizing noice and smoothing image
Gussian : control amount of bluring
Highpass : enhansing high-frequence like edge and make image appear sharper
Nagative : make bright pixel dark , dark pixel bright and enhansing image details
the enhancement increase the contrast distribution to increase the quality and visaldzation , so we apply histogram to increase Contrast in image by distrbuting the grey level in the image .
After that use segmentation techniges to extract the features of ROI by used sobel oprator, simple it detecting edges by highlight areas of rapid intensity transitions.
then used morphological opration to analyse and mainpulate the shape and structaure of objects with in the image by two fundamental :
1-Dilation : to fille the holes and increasing the Size of objective
2-Erosion : to remove Small objective and Smoothing object boundaries
the Classification step depend on support vector machine(SVM)that takes extract feature from image (ROI) and put it as input for SVM algorithm . then it works by finding the hyperplane that separates the different classes in the feature space. after Comparing the actual data and the expected data, the result is extracted as Cancerous or not Cancerous.