oterem / moleagnose

our new code repository - https://github.com/Oterem/MoleAgnoseFinal

moleAgnose -An application for detecting and diagnosing suspicious moles on human skin.

This appliction is being developed as part of 4th year final project at the software engeneering department at Azrieli College of Engineering, jerusalem, Israel.

This project is developed as native android application using open source computer vision library - OpenCV

The project is advisored by our professor Dr. Assaf Spanier.

Tools & Practices

We're trying to use the best tools and practices in order to produce a quality engineered product.

Category	Status
Version Control System	git & github
Recommended IDE	Android Studio
License
Code Style
Issues
Project Management Board	here
Documnetation
Project diary

Software Design

Architecture
Class diagram
State charts
Sequence diagram

moleagnose's Issues

add histogram option

calculate euclidean distance between seed point and skin point

Extract mole features

Build notification to user

extract image into opencv "imshow" function

need to take the picture and import it to opencv function and start analyze the image

reduce noise using filters

run the analyze function in a different thread

seperate segment and feature extraction into two seperate threads

Create another AsyncTask

switching from camera intent into camera object (camera2)

detect human skin

calculate "roundess" of contours

1. find minimun enclosing circle.
2. comare this circle with matchShape.
3. generate numeric result.

set the creaeChooser intent to browse directly to the relevant path

determine automatic threshold

sample 4 corners of picture and set a avarage color for threshold

Add crop image option to camera

segment the mole in full color

bitwise_and the binary image with original colored image

segment blob based on region growing algorithm

Support display of high resolution images

The "regular" Bitmap object can't display high resolution images due to memmort allocation issue. Need to convert to Bitmap factory and resolve the problem

quantify assymetry of contour

Etablish DB

segment the melanoma

after setting threshold, need to segment the melanoma using contours

sample points for blob and skin

Using imageview onClickListener need to sample 2 points

implement region growing based on two sampled points