FYP - AR indoor navigation

complete the test  angle: Trial 90° 60° 30° 1 87.2 61.1 31.2 2 93 59.3 29.8 3 88.3 60.4 30.5 4 91.3 60.8 30.9 5 89.6 58.9 28.9 6 90.5 61.5 31 7 88.7 59.7 29.6 8 93.8 60.2 30.2 9 88.9 62.9 30.8 10 89.7 59.6 29.9 straight: Trial 5m measured 10m measured 1 4.84 10.02 2 5.09 10.08 3 4.95 10.08 4 5.03 10.1 5 4.97 9.8 6 4.96 10.06 7 5.05 9.92 8 4.9 10.12 9 5.01 9.98 10 4.59 10.04 straight with turning: Trial 5m measured 10m measured 1 5.72 10.88 2 5.84 10.94 3 5.69 10.72 4 5.76 10.25 5 5.78 10.88 6 5.66 10.67 7 5.88 10.97 8 5.74 10.79 9 5.82 10.85 10 5.53 10.73 Also recording the building information(routes) of the 6th floor of the EE building And I've done a new poster Meeting: No meeting this week

 For poster and report I need some data to support the navigation precision. Test list: without turning: 1m, 5m, 10m with turning: 1m, 5m, 10m turning angle: 90°, 60°, 30° and make them into a graph (error %) this week I've done part of the test  1m for straight and 1m for straight with turning 1m straight 1m with turning (+0.63) 1.10 1.68 1.00 1.61 0.98 1.72 0.97 1.59 0.91 1.66 0.95 1.61 1.02 1.70 1.03 1.71 1.00 1.63 1.08 1.59 And also I've done a draft of poster   Meeting: poster has too much words which make it don't like a poster Poster: 1. more graphs 2. less words 3.larger the size of words

 this week I start to transfer the focus from project to bench, and the development works gradually ceased. The problem is how to let the arrow point to a right direction. Solution: Poster: 1. 2 graph of how the system works and the flow chart of it 2. Done the text works of the first part (intro) Meeting: Solution works well and as expected. And the poster figures or graphs I've created is good but too technical, I need some simple graphs to show people what I've done 1. need to attract people by words 2. story of your project is important 3. think about any benifits: reduce asks/ problems      reduce the queue

 Adding a segment/ full mode to demonstrate the part/ full of the route, and also adding a switcher to change the AR navigation ( route) or only arrow based navigation. Segment/ full mode works well and deliverable, But the arrow mode is not good, the arrow works well at sometimes, and is may also point opposite even if the destination is at the front. Solution: at some crossing point or some check point at the crossing. and make them point to the crossing and detect whether need to point to the next one. Meeting: update is well, but the arrow mode problem should be solved.  And the focus should be on the bench inspection instead of full on the project Bench should start with the problem no the tech and offer a solution So, I need to prepare a story of the problem, and give out the reasons why my solution is good by comparing with the others.

 The application suddenly have some bugs, and I fixed it. Problem: 1. fixed length step (72 cm) used to using sensor to calculate how many step admin has walked solution: I see the "measure" application in iOS system is good, and which can get the distance with a high precision, so base on this, I think I can using the similar method or APIs to enhance the precision of my project. By some research which is using AR kit, so which is similar with the navigation. So this week I've done a lot of tryings and testing to make sure this new method is available. Meeting: demo is good as well. Problem: 1. precision is not good. reason: because QR code is not fixed at the place at present, as also the direction of user or the initial state is not fixed which causing the error occur. So, I may say this is acceptable for present. 2. Arrow can be more simple to give a clear instructions (also trying to reduce the bad feeling by the not good precision) solution: 1. simplification the ar...

 Because the absence of the meeting last week, I don't have some significant updates just optimize the algorithm. Meeting: updates during the holidays are very good. but have some problems problems: 1. can it be a website-base application again?  this is possible, but not sure whether the time is enough for me to  2. need to be recalculated or restored by a better route? yes 3.the recording method is not good, by using fixed length step to accumulate the distance  this can be fixed by changing the recording sensor