Future Improvements of Implementation
From SoftwarePractice.org
Improvements
Our program is not perfect, and nor are we claiming it to be perfect. In fact, there are numerous improvements that can probably be considered when implementing the program in the future.
Improvements include:
- Provide a function which calibrates the image so the width of the black and white lines are irrelevant. This can ensure a wider range of barcodes to be processed as we are able to overcome this limitation. Since the initialisation and termination codes are known and easily read, it is a suitable target to calibrate the read image function.
- Further research into allowing more image types. Currently our program is limited to processing .jpg images. While we are fairly sure that it is able to handle other image formats, as mentioned in the limitations section, we have not tested these boundaries. In addition image encoded by .gif are unable to be handled by our program for reasons which we are unsure. To improve the program we would need to be sure that the system is able to handle a wide range of image formats.
- Read tolerance within the read image function to overcome images containing pixel error. It was discovered within the program that despite deblurring and denoising processes there remained some pixel distortion of the image. At the moment there does not exist any further filter process which is able to handle noise which remains after the currently placed filters.
To rectify this problem we could establish some threshold variables so that as long as the image contains a certain percentage of black to white, or white to black, it will be interpreted as such. For example, a threshold variable of 80% could be inserted so that as long as the segment contains more than 80% of a particular colour it will be interpreted correctly.
