After experimenting with different types of trees in different scales both with and without the campus layer on, I found it is probably the best to show detailed, relatively large size trees without campus layer.
Instead of moving the marker to show the complete road, I've animated the road in motion, in that case, object can have large scale with more details without worrying some parts may not be seen if the marker is moved out of camera view.
Problems: The lighting condition is not desirable in the red centre lab, it is too bright and uneven. However, the 'trees and road' layer did seem to appear on campus layer relatively well despite the bouncing of the model due to the marker not being flat on ground.
Next step: Finish constructing all the trees and arrange them into one layer with campus road .
The original tree I was using has too many polygons, and it has taken a lot of time to place them onto the right places, so I replaced it with a low polygon tree found in google warehouse.
Trees lost their texture and all became black after imported into 3dsmax and viewed with BuildAR. I think this is probably due to the resizing of model from its original scale, the amount of detail is not visible in such a small size.
At this stage, I have successfully imported the Cad file of the campus into 3DS Max, which didn't take much time as far as the level I'm concerning. I created a simple marker for the campus and tried out in BuildAR. The result is satisfactory--the texture is right, the orientation is correct, the added text is appearing. However, The embedded animation is not appearing.
This rendered animation shows the building (red center in this case) being highlighted.
In this video, I had Assigned the highlighted building into another marker and showed it when needed. However, I still couldn't get animation to work like in 3ds Max. Anyone out there know how to do it? Also, it is very hard to put the highlighted building marker on the right spot, it probably needs to be adjusted both in 3ds max and buildAR
Things to do: 1. Make a solid base for making it easier to put everything on it in right position. 2. Continue modeling the selected buildings using different techniques (I'm still thinking if the photosynth reconstruction will really work, or I should just use the model and add more details into it) 3. Collect important information of each faculty and manage to put them into animation (display the information) 4. The campus model itself does not have any plantation, I think it will be nice to model the trees and show environmental aspects of the campus 5. Create different markers and arrange them into a logical way
The highlighted blue area is UNSW campus, which will be one of the fundamental components in my project, and the red highlighted spots are my selected buildings which I will be reconstructing. I was fortunate to get the base model from Jeremy, so my focus will be on how to represent the UNSW campus guide in a creative and logical way.
Buildings to be reconstructed:
Round House, Law Library,Chemical building,Red Centre, Quadrangle buildings and Main library.
Project Proposal In this project, you will be exploring the possibilities of multimedia presentation, which will be fully engaged with the audiences. Essentially, the project is aim to produce an interactive 3D campus guide which will establish the knowledge of the campus (direction of the destinations, features of different architectures, building history, faculty facilities and etc. ) Basically, There will be a marker with the assigned 3D model of basic campus landscape, and several small markers representing the important buildings in the campus, in this case will be photogrametrically reconstructed (using modeling skills to help to refine the model), and then animating the add-on information in 3ds Max to enrich the experience of getting to know the campus. Ideally, there be a trigger relationship between the landscape of the campus and selected building, which means when the building is on the right spot, the embedded animation will be triggered otherwise not. Unfortunately, buildAR is not able to achieve the goal, and if it needs to happen, software LinceoVR needs to be studied.
The alternative way of achieving this is to having 4 sets of markers, first is the basic campus landscape, second is the building names, third is the highlighted spot of the building and direction, forth is the actual building and the animated building information.
Assessment Criteria
Have thought and care been invested into all aspect of the design process in context to the technology you have experimented throughout the course, including a noticeable attempt to engage with the technology in a logical and creative way.
Blog:
Is your blog laid out and easy to follow, with all task requirements posted to your blog?
Image:
Are your images of an acceptable quality, showing thought and consideration to image selection and camera framing? Are all images not your own referenced correctly?
Text:
Is any text you have written concise and informative? Is the text significant to your project, outlining aspects of your process that may not be immediately apparent through visual material?
Installation:
Does your installation work? Is there a novel approach to the interaction between real life and digital content?
Marker The marker is designed according to the section cut of a particular scene, since the Esherick house has a simple geometric shape and it would be consistent to keep the simplicity, although the building itself is full of details.
Esherick House located in the Chestnut Hill neighborhood of Philadelphia, Pennsylvania, USA. The hose is designed by American architect Louis Khan, constructed from 1959-1961,I built the model in sketchup and will be using it for this assignment.
I will be incorporating both 3D interior views and animations to draw significant aspects of the building. One of the animation effect I will be using is lighting, The unique programming of the building and elegant choice of the materials have enriched the atmosphere of the building interior, accommodating with Kahn's design of built-in fixtures, the house becomes a delightful living space.
Testing Animation of lighting
Testing 3Dobject and Marker relationship in BuildAR
This is how I positioned my model in sketchup according to the axial, The model is in scale of 1:50.I made a base same size as the side I wanted to show, and put the marker in the middle, so what I expected to view in BuildAR is the model fit on the base perfectly. However, it didn't turn out to be the right size.
So apparently, the size of the model appears on BuildAR needs to be adjusted both in sketchup and BuildAR, problem solved!I carried on and tried to exported my sketchup model as obj file and imported into 3dsMax, it didn't go through. Then I tried to use 3ds file, it worked, but I lost all the textures. I accidentally saw a youtube video about transfer file from sketchup to 3dsMax, it used the sketchup file without exporting, but it only works with sketchup7
And it works with 3dsMax perfectly since the file contains all the information.
I tried the animation on my model and below is a glance of the result, the sun path still needs to be adjusted,but the idea kinda works.
This video basically shows what we will be achieving for second assignment--Create 6 different markers, Assign the building interior into these markers and then record them live in BuildAR.
Tissot has made Augmented Reality into their online shopping, consumers will be able to try on a selection of styles and select the finish and colour of their wristband from the Touch collection. It opens a new way of e-shopping and brings more interactive and also maximize the customer satisfaction.
Installation 1Object of Interest ‐ Photogrammetry and 3D Printing
Concept When the world is unified, every building is identical, every inch of concrete is equally modified.... Where will civilization exits? This piece of object evokes the idea of museum as core of cultural revolution leading the human civilization in an emotive and artistic gesture. The base of the object built up with equal size cubes represent a standardized civil life which is imaginably decaying in designer's utopia dream, where the boring 'contemporary city' is upcoming in a bigger wave.
Comparison of Real object and 3D reconstruction (Texture image from voxels)
Comparison of Real object and 3D reconstruction (Projection mapping)
Inspirations
It was initially confusing to understand 'Altered object' since the physical presence of the object in someway blocked me thinking freely. I decided to use the Guggenheim museum model as the base model since the result in 3DS MAX had turned out the best.
Inception has provoked my initial idea in this project, architect created a mobius-like band, shattering reality and challenging the possibility of an imagined dream world.
This bending moment is extremely exciting for an architect, where the deconstruction of architecture has achieved perfectly
In the limbo state, the architectural Utopia is manifested here. Le Corbusier’s never achieved remarkable Contemporary City
I was able to get 100%synty for this set of images, which means they have matched pretty well. However the point cloud I got from it is not so great, many details are missing.
I put it through the meshlab to see the reconstruction result
The result is not very satisfactory, too many noise, many details are missing as well. So I decided to take another building which hopefully will have a less consistent pattern on its facade.
The process of photosynth went pretty smooth, 100% synthy. I was also able to get more point clouds than previous building.
Although it still has lots of noises in meshlab, I was able to get more details.
This is the mesh created with texture, since there are too many noises in front of the building, I wasn't able to create a mesh wrapping to back of the building face, which means my mesh is like a closed ball, you have to zoom in to see the building face.
I tried to import my .obj file to 3ds Max, and this window popped out saying 'nothing found to import' :(
The first set of photos hasn't been very successful, 96% synthy
Selected view of the model
Camera Information Canon EOS 400D Digital SLR Lense: 18-55mm
Camera setting f: 5.6, Iso: 400, shutter speed: 1/200 Back of the model is shallow and lost details which makes it harder to match with images in different angle.
First result of surface reconstruction. pmvs parametric setting: level: 2 csize: 2 threshold: 0.7 wsize: 7 minImageNum: 3
Since the reconstruction result wasn't satisfactory, I had decided to retake another set of photos
