You are on page 1of 4

Pulse Monitor

Recording your every heart beat for a healthier life


Instructor: Ghassan Hamarneh

adfa, p. 1, 2011.
Springer-Verlag Berlin Heidelberg 2011

Re-cap of Project Goal

Pulse Monitor is a camera based IOS application, which can record the user's heart rate and
analyze it based on personal information. With different measurement options implemented, user can
choose to record heart rate during rest, before exercise and after exercise. The heart rate information
will be presented as ECG (electrocardiogram) on screen. In addition, all records will be stored on
device in order to let users view analysis reports.

Milestones and accomplishments

Since we are developing this application based on our original project plan and timeline, all
processes expected by March 15th have been finished. Below is the list of detailed milestones and
finished steps:
ID
1
2
3
4
5
6
7

Task Name
Project Proposal
Requirement Document
Design Document
Database Design
User Profile Setting
Calendar and Notification
Measurement Algorithm

Finish Date
2/15/2015
2/20/2015
2/26/2015

contributions of each member:

Responsibility of each member has been discussed on project proposal, as agreed; requirement
document and design document have been created by the whole team at Feb 20th and Feb 26th.
Xiang Zheng:
Xiang has designed and created application database, which will store application operation data and
user information. User profile setting feature and interface has also been developed in order to
communication with backend database.
Peijiang Miao:
Peijiang also contributed on developing User profile setting feature and interface. In addition, he
designed, developed and tested calendar and notification feature as well as measurement algorithm.
Tommy Du:
Tommy contributed on all the user interface design and implementation. He also performed feature
integration in order to test User profile setting feature and notification feature together. He created
detailed test plan and test cases.

Self-Evaluated Progress

The progress achieved from Feb 15 to March 15 had been on schedule as originally planned.
The User Profile setting component of the application took slightly longer to accomplish as we were
still getting familiar with xcode.

Duties of Each Team Member

Tommy Du will be responsible for developing the frontend of the application, which includes UI
design, report visualization and configuration panel.
Peijiang Miao will mainly focus on the backend design and development, which includes API
development and algorithm research.
Xiang Zheng will be in charge of raw data parsing, data storage and data retrieval.

Anticipate Problems

Having little knowledge on Swift development might hinder our project progress.
An absence of team member(s) at a critical time due to illness or other unforeseen circumstances can
bring possible incompletion of the project.
Implementation of ECG generation algorithm may be challenging.

Material

Data: Data will be obtained by using a pulse oximeter which illuminates the skin and measures
changes in light absorption, which is a process known as photoplethysmogram (PPG). This will be
achieved by using cellphones camera and flash light.
Software: Xcode
Tools: Microsoft Office, OpenCV, Github, UIKit, AVFoundation and CoreGraphics.
Algorithm: By using camera to capture the video with high resolution, each frame in the video will
be analyzed by getting RGB value of every pixel. Before processing, raw data will be filtered by
using Fourier Filter algorithm. Then preset variable will be used to amplify the color value linearly
within proper time scale. After amplification, fast Fourier transformation will be applied so that
instant waveform will be presented [2].

Detailed anticipated results

The following result will be produced:


A mid-product video will be generated during processing. The video resolution will be appropriately
designed as 640 x 480 with 60fps in order to take both accuracy and efficiency.
The final result will be in the form of ECG. Each heart beat will be output as a pulse on the graph.
Values on the waveform will come from the average RGB values of each frame and be filtered by
Fourier filtering method. After filtering, values will be amplified so that small changed can be
detected [3].

Software input/output: the application will take standard video sequence as input and ECG graph as
output.
GUI Sketch:

Reference

[1] CSAIL. MIT. Video Magnification http://people.csail.mit.edu/mrub/vidmag/#people,


[Assessed: Feb. 5, 2015]
[2] Ghassan Hamarneh Project Guideline http://www.cs.sfu.ca/~hamarneh/courses/projectguidelines-340.html, [Assessed: Feb. 5, 2015]
[3] Neal Wadhwa, Michael Rubinstein. Phase-Based Video Motion Processing
http://people.csail.mit.edu/mrub/papers/phasevid-siggraph13.pdf, [Assessed: Feb. 5, 2015]

You might also like