This week we tried to finish our model and send it to the printer while working on our final report and presentation. We started the model over due to the fact that we could not figure out how to stitch the facades together. Maria A. uploaded 70 of the best pictures into 123DCatch again which took time to process each facade. This time, when we uploaded each facade into MeshMixer, we removed all of the extraneous parts, filled in the holes and smoothed out the surfaces before we added another facade into the program. We were better able to stitch them together with increased knowledge of the software. This time the facades were all stitched together. The only part we had trouble with were the 2 roofs. LeBow has one high roof and the highest part of the building and another one that slants downward in the lower section of the building. This was harder to artificially create in MeshMixer. But we did it and smoothed out any last minute imperfections.
Overall this program was very difficult to use. MeshMixer is a software that is supposed to be utilized for small inanimate objects and not for monumental architecture. Had we finished image collecting earlier with our phones, we would have had more time time to either become familiar with the software or switch to a different software and try alternative methods. Below is an images of our final deliverable and what it will look like when printed.
In week 9, we were slightly off schedule as we still did not have our model ready to be printed. There were some issues with 123DCatch and Meshmixer that we encountered. 123DCatch was very straightforward to use. We just had to upload 70 of the best images that we had of each facade. Because 123DCatch could not take zoomed in images, we could not upload any of the images obtained by the drone. The images from our phone were great quality and translated well in 123DCatch. But it was very difficult to upload in Meshmixer because 123DCatch creates .stl files whereas MeshMixer uses .obj files..So we had to convert them. Once we did that, the facades lost a lot of detail. The concrete sections of LeBow rendered well but the glass sections of the facade did not and gave a lumpy and "melting" effect. For week 10, we plan to stitch all the facades together, smooth all of the surfaces and add the roofs of the building. Below is an example of what we were able to create in MeshMixer.
It is still not completely stitched together which we need to work on next week. We tried to use the zipper tool in MeshMixer to stitch it together but so far it has been crashing every time we try so we are going to have to look for tutorials online to find another way to stitch the model together.
This week, we have been modeling LeBow in 123DCatch. We realized that 123DCatch cannot recognize images up close so we had to capture images from far away. 123DCatch also can only take up to 70 images so we uploaded 70 images of this facade to create it. Below are images that we collected across Disque and were imported to 123DCatch and used to create the back facade.
18 images of LeBow
Model in 123DCatch
We finished the first two facades completely in 123DCatch. All we have left to do is upload the pictures for the other two facades and stitch them in 123DCatch. Once we have all the facades created in 123DCatch, then we will upload the model into MeshMixer and fine-tune any imperfections and delete any unnecessary parts that are still in the model. Next week, we have to finalize our model as we are presenting next week.
Last week, we called Drexel Facilities in hopes of being able to get access to the roofs of other buildings so we could obtain the best images but they denied access. So, we made our last attempt capturing images with the drone.After not being able to capture many images, we went to several different buildings and captured images with our smartphones. Maria and Sena went to the highest floors in Pearlstein, the Papadakis Integrated Sciences Building, and Korman and captured images. Maria L. and I went to the highest floor in Disque and captured images with our phones. We met up once more outside to get the rest of the pictures we needed mostly from the facade across University Crossings. This week we are in the process of 3D modeling in 123DCatch and obtaining any other photos that we need.
In lab this week, we mostly continued according to our plan. Sena has become our expert flyer as she was able to fly the quadcopter up to the middle section of LeBow. We had some complications in lab this week. This sprinklers turned on when we were flying our drone, but luckily it was unaffected. Then, our drone was stuck in a tree soon after. Because of these complications, we did not acquire the best images of LeBow from the drone. I was able to capture images with my phone of the higher parts of LeBow by going to the highest floor in Disque.
We know that we will be unable to reach the top of LeBow so we will rely on going to other buildings and using our on cameras and smartphones to capture the images of the tallest parts of the facade. Our plan is to go to other buildings and get permission to go on top of the roof or go the the highest floor to take pictures using both our phones and the drone if possible. Most likely, we will go to University Crossings and obtain the facade facing Market Street. We will go to Disque to obtain images of the facade facing Disque. Lastly we will go to Pearlstein to obtain the high part of the side facing this building.
Last week we attempted to take pictures of the smallest facade which is the side facing the Paul Peck building. Too many people were near the area so it became hazardous to fly the drone in that area. We moved towards the quad, in the grassed area, so that in case the drone fell, it would have a softer landing. As Maria and Sena continued flying, many people approached us on comments about the drone with their own remarks. One person who had previously bought this item told us that there was no use flying it outside because even the faintest breeze could way it in any direction. We were already seeing problem occurring in previous weeks and his own results supported what we started to notice. So we adapted and paused any time there was medium to heavy wind and resumed if there was light to no wind. Both Maria and Sena became more proficient in flying.
Even in our best attempts at flying the drone, we couldn't control certain parts of the drone such as direction and landing. We soon realized that it is going to be nearly impossible to fly the drone to the top of LeBow. Even the person who had experience with the drone said to not attempt it at all. From this experience, we decided to obtain as many pictures as we can with the drone and obtain pictures with cameras within our phones for the higher parts of the facade. Considering that LeBow is the tallest building on Drexel's campus, is is going to be a daunting task to obtain the images up top.
Outside of class, we also noticed the picture quality of the drone versus my phone.
Capture from GV3
Capture from Samsung Galaxy S6
Zoomed in, we can see that both image qualities are great but there are certain advantages and disadvantages of both. The GV3 can obtain good quality pictures at great heights but it is difficult to fly. It is hard to fly both very high above ground and at a stable level. My phone can obtain better quality pictures even when zoomed in to the max capacity but it will be difficult to obtain images with my phone towards the middle and top sections of LeBow.
Last week in class it was too windy to perform any flight plans so we did not even attempt it. Instead we tried to see if we could connect the controller to our quadcopter inside and do some light flyng outside for practice. Everytime we raised the throttle, the drone would fly even at the slightest touch. But when we would release the throttle, our drone would lose connection and would have to be reconnected, proving it to also be very inefficient. So we decided to fly at least next week in at least better weather conditions. Below is a picture of the best image that we obtained in lab.
First Image of Gerri C. LeBow Hall
Instead of flying, we gathered measurements of all facades of the building. We used measuring tape and alternated positions to acquire the lengths of each facade.
Measurements:
The entrance facade: 53'6"
side facing Disque: 196'9"
side facing Bossone: 31' 9"
side facing Market St.: 173' 6"
.jpg)
