A Prescriptive Tillte

Only available on StudyMode
  • Download(s) : 90
  • Published : April 7, 2013
Open Document
Text Preview
Image Changes Detection (ICD)

SRD

Source Require Document

August 20, 2012

Index

Index- 2 -
Introduction- 4 -
Feature list- 4 -
Library dll- 4 -

Window form test application feature list and definitions- 4 -

General required definitions- 5 -

Useful links- 5 -

Implementation:- 7 -
Requirements- 8 -
Implementation plan- 8 -
Step A: Payment (None) – Done- 8 -

Testing final delivery: Payment (50%)- 8 -

Deliveries- 9 -
Pages layout- 9 -

Purpose

This document contains the requirements of Image Changes Detection (ICD), features list and high level architecture.

Introduction

Image Changes Detection (ICD) is a .Net windows library to provide and manage images deference's, by compare between 2 images. It based on the implementation on codeproject (OpenCV[1][2]), and it will be a .Net wrapper library. In case the performance is not good enoght then we will use the OpenCV.

Feature list

Library dll

This is a .Net 4.0 C# library which wraps OpenCV/ and provides same abilities in addition to sending images than video.

1. Name ImageDiffDetectionManager.dll

2. Low CPU (be sure CPU will not be high @ least same as the demo in that article)

3. Implement the interface iImageDiffDetectionManager (see interface in the end of this document)

Window form test application feature list and definitions

This is a graphic window form application to provide a test application to use all the abilities provided by the ImageDiffDetectionManager.dll library.

1. It should be compiled as Any CPU

2. C# .net 4 framework

3. The form include

a. 2 radio bottoms

i. Draw Rectangle: get and display only rectangles

ii. Draw Bitmap: get and display to Result image only bitmaps pieces on its corresponding location (in Source image), list of crop part images include the changes

b. In addition, the form include

i. Source image: static image, load from disk (browse button), which can be any resolution

ii. Container (any such as frame): when select draw bitmap option, it will include the list of images.

c. go button, when click on it: the text changed to stop, and start loop:

i. Save current image as source image

ii. New image = Draw random shape (line, pixel, circle), random location, random color, random size/length

iii. Send new image, and current image (section i) to low level to process

iv. Display the new image to Form and draw the rectangle around the changes, if rectangle inside another draw only the bigger one

v. Save the new image (without rectangles) as current image , with last shapes

d. Click on stop to stop the loop.

e. If the user select option 2, the loop will be:

i. Save current image as source image

ii. New image = Draw random shape (line, pixel, circle), random location, random color, random size/length

iii. Send new image, and current image (section i) to low level to process

iv. Display the list of changes cropped images

v. Save the new image as current image , with last shapes

General required definitions

1. You can't provide 3rd party bin files, mean you need to include all 3rd party source code and compile together with our solution/project

2. Maximum image resolution 2048×1536

3. Compare 2 images with 30% changes should take less than 500mili sec when compare maximum resolution

Useful links

1....
tracking img