README: ImageEnhancer

ImageEnhancerExample, version created by Marcus Nagy for course ET2584 at BTH.

How-to-Use

At load-up there is one button available named Load new image, this button allows the user to selected an image from the file storage on the device. After selecting image 3 new things appear.

• Radio Group: Test Enhancer - Choose this to prompt for 4 different actions of image enhancing. V-Transform - Choose this for using the V-Transform to enhance the image; when selected and click on button Modify image the user will be prompted for selecting number of segments. Smart Enhance - Choose this to enhance the image in a smart way, where the best value in Median Absolute Deviation for different segmentation values of the V-Transform will be returned.

Note: Smart Enhance requires substantial computing power and could slow down your device, recommended to be used on newer devices only.

• Modify Image: This button on click might prompt the user for more input data, depending on what enhancer is selected from the Radio Group. Once the enhancing is executing there will be a progressDialog displayed giving feedback for the progress.
• Save: Button will prompt the user for text input that will require the name for the enhanced image being saved to pictures on the device.

Note: The button Save might only work properly on API 29 or above, and will display a Toast if you're not able to use the function.

• Info: Button will display informative information about segmentation by creating a Toast. If no method using segmentation is used a message saying "Nothing to display" will show.

Note: This information can be modified to display other things for other image enhancing methods.

To Hide/Show the options and buttons in the bottom "tap" the afterImageView.

To Hide/Show the Load new image button '"tap" the beforeImageView

Note: Using any of the available image enhancing methods will make a Toast appear when Info is clicked, displaying information according to the function selected. When V-Transform is selected number of segments and the actual number of segments calculated by the algorithm will be displayed as "Chosen:" and "Actual:", which finds the closest factor to the selected number of segments; Smart Enhance will display the best number of segments. Lastly if nothing has been explicitly modified and any other image enhancing method is selected the Info button when clicked will tell us "Nothing to display"

Version changes

New variables found in the MainActivity.java will have a comment in the style of //NEW: (details). All new methods deemed necessary for explanation will have comments above them in the style of /** */ .

*Information about new classes and their function can be found in next section.

New classes

There is a couple new classes to allow for V-transform and Smart Enhance.

• Pixel: Class is a representation of a pixel with x- and y-cords and a value. Implements Comparable<Pixel> and allows effective sorting on the Pixel value. Class also uses Factory Pattern for easy way of creation, since many pixels are often created.
• TemplateEnhancer: Abstract class that contains logic/methods which both of the V-Transform and the Smart Enhance uses. Implements ImageEnhancer and is constructed after Template method Pattern to be a template for both V-Transform and Smart Enhance.
• VEnhancer: Class extends TemplateEnhancer and performs the V-Transform on a given image with specified number of segments.
• SmartVEnhancer: Class extends TemplateEnhancer and performs the V-Transform on all possible factors but returns the image which produces the best Median Absolute Deviation value (which is the value closest to 0.25)

Changes from Feedback

There have been some minor changes to improve layout for multiple devices by adding an Info button and removing a textField. This allows for older APIs to be used. Save button now works properly but also with improved feedback to the user.

Note: Changes are visible in How-to-Use and everything old has been removed.