RIT Internship

The morning meeting was short, we were missing a lot of people. Emily and I explained the Archimedes codex. We decided to focus on the outline and getting that done. We spent a lot of time researching the Ambrosiana Library and then the Jubilees Palimpsest. We had to figure out what they actually are and the different texts that make up the palimpsest. We are starting to think about our final presentation, we have half of it pretty much done from the presentation that we had to make for Dave. I feel like we have a good grasp on our topics.

  Outline   Intro Background: We are working to uncover illegible text from the "Jubilees Palimpsest" and find corrections and later additions on the Selden Map of China     Purpose/Technique: We use hyper-spectral imaging (several narrow spectral hundred bands) and multi-spectral imagining (fewer than 50 bands) The images are digitally processed and combined to create images with more characters that are legible than those in the regular RGB image Processing tools include Principle Components Analysis (PCA), Spectral Angle Mapper (SAM), Spectral Profiles Selden Map Background: The Selden Map dates back to the early 17th Century during the Ming Dynasty The map includes approximately 15 countries and shows a system of navigational routes China, Borneo, Japan, Korea, and Taiwan are five countries with the most significant results Focus/Results: Green corrections in Borneo, Korea, Taiwan, and Rivers of North China Missing texts and nam

Today was no fun because there was no morning meeting. My day started with copying over more data files. Than I had to make cubes for the new data. At 6 we left for the observatory and the bus ride was fun. We listened to a presentation and learned a lot about the instruments they use there. We didn't get to see anything though because it was too cloudy. It is late now so this is all I am writing.

The morning meeting was fine- we were told to find out what documents we are actually looking at. In the morning Ashley, Peter, Emily and I helped out with Titus and Aditi's lab by wearing their eye tracking glasses around outside. Then we came inside and made more cubes of the images both the narrowband and the fluorescence. Then I ran PCA and SAM scans on the images to try and pull out any under text.

The morning meeting was short, we went over our last abstract. We started off the day getting new data from Roger. We spent a while trying to figure out how to make cubes. We asked David Lewis for help a lot and he didn't mind too much. We pretty much spent our day learning how to make data cubes, calibrate them and rename them. We did help Ashley and Peter in the 3D printing and that was cool. So that was my day.

Today I missed the morning meeting because I got trapped in the monsoon. When I got her Emily and I began talking with Roger about what we would be doing from now on. We started by trying to figure out how to copy the data onto our own accounts. Then Emily and I tried to understand all of the different images that were collected, the difference being the filters and the wavelengths. Emily and I started playing around with the images and ran SAM and PCA. I tried to pull out the undertext as much as possible.

Today the meeting was short. It ended up being a very interesting day. We started off running anomaly tests on all of the SWIR data, but we didn't find anything. Then I ran anomaly tests on some of the VNIR images. Chris was in the reading room again but he didn't talk much. For lunch most of the interns went outside for lunch and we had a lot of interesting conversations. We continued with our anomaly tests until we went to visit Titus and Aditi. They ended up asking us to help with their lab so they started pushing around Emily in a chair around the hallway with the special glasses on. Then it was Peter's turn so Emily and I made peter a special hat that said I <3 Imaging Science, to keep the sun from interfering with the glasses. But the software wasn't working so they made some of us do a virtual reality experiment. Overall it was a pretty fun day!

The morning meeting was pretty short, we finished up the abstracts and people didn't have a lot of bad things to say about ours. We spent the morning trying to find people that read Chinese and went to all four floors. We had a fun time trying to print out some of the images we had of both the Borneo and Taiwan port. Everyone we asked thought that it probably isn't any Chinese characters. Ronny and Gerry helped us tried to help us out with the Chinese and gave us some other creative thoughts as to what they might be. Ryan was no help, which was disappointing, but whatever. After we gave up on the Chinese writing, we looked back at the presentation. I decided to go back and redo the SAM images of the red flowers in Japan, both the smaller images and the entire map. Then I went back through the entire presentation and just fixed little things. Today Ashley, Peter and Tristan came up for lunch and they had many interesting conversations. Now I am stuck here with Emily ;)

We once again started the day with going over other intern's abstracts, doing about four today. I started with finishing the map of SAM overlay images comparing the three green pigments. Once I finished the green pigment SAMs I began working on the blue pigment SAMs. I pulled a blue pigment from the blue outline on an island off of Western Sumatra. I ran all of the VNIR images with SAM for the blue pigment and then did the interactive stretching to help bring out the similar pigments. I then put all of the images together to make an image of the complete map (minus one section that I cannot find). I added both of the maps to the presentation and now I think we are up to 63 slides.

The meeting was a little bit longer today. We read through four of the abstracts and suggested some edits. Once again we began working on the presentation. I continued with running SAMs for the three different pigments of green and making sure that we had the correct bands for all of our images in the presentation. Emily went back through the slides and added short explanations as to what the data represents and means. We decided to download all of the VNIR and SWIR files from the hard drive so we have everything that we needed for today and tomorrow. Tomorrow we will just continue with the SAMS, but this time with more focus on the red and blue pigments. We will begin looking at the SWIR files and see if there is anything interesting there.

First we had a short morning meeting with Joe and we found out that we will begin presenting our abstracts tomorrow. Emily and I went over some of the edits that she made on Friday to the presentation and then we began looking at the changes that we still needed to make. We focused on remaking most of the plots and finding the exact locations in which we pulled the pigment samples. We made sure to label everything correctly and make it clear as to what we did. Once I made the individual graphs for all the green pigments in each different location, I combined them into one large graph. With this large graph, I grouped similar pigments into three different groups and began to run SAM images to get a clearer picture of approximately where each different pigment group exists. There is still a lot of work to be done on the presentation for tomorrow!

In Document Restoration our main focus is analyzing documents through pigment analysis and image enhancements. We are currently working on finding corrections on the Selden Map of China created during the Ming Dynasty in the 1600s. We work with illegible texts that have either been erased, covered up or faded. We use hyper-spectral imaging which consists of several hundred color bands, and multi-spectral imaging which consists of less than 50 color bands. The spectral imaging allows for a more precise and detailed image than normal red, green and blue (RGB) images, making it more legible and visible to the human eye. We use techniques such as Principal Component Analysis (PCA), Spectral Angle Mapper (SAM) and Mahalanobis Distance. These techniques show inconsistencies in the paint and identify where different pigments are located, which could point to a later addition or co rrection. These same principles are applied when working with illegible written text, typically being palimpsests

At the morning meeting we watched another short video with Matt. Then Emily and I began to look at the new data in VNIR1 and VNIR5. First I focused on West Banda. I used SAM to compare the blue pigments that I had plotted on the graphs to maybe see if there was actually a correction made. Then I collected several PCA images to try and identify the writing inside the port. When we got as clear of a picture as we could of the writing in the Banda port, Emily and I conducted an intense search of the Chinese language to identify the symbols. But that didn't turn up anything that helpful so we began organizing all of the data that we have collected this far that could be helpful to Robert/Bob. Then we moved on to analyzing the pigments in the lower corners of the map. 

First we had a short meeting with Matt and he showed us a video. Then I started to plot the various greens throughout the map. Then Emily and I learned how to enhance images, so we began doing that for the writing that we found in Taiwan. We decided to zoom in on the enhanced images and print them out so we could try to make out what was once written there. We cut out what appeared to be images and then taped them onto a larger piece of paper in the correct arrangement. With Dave we traced over our masterpiece on the computer and it did not look like writing but at least we tried. After that we transferred new images to our computer to begin addressing the other questions. Finally, we began to write our abstract!!!!!!

First we had a short morning meeting. Then Emily and I learned a new enhancement process called high pass on ENVI. We focused on the circle and writing in Taiwan, trying to make out what is written inside the circle. I saved many of the images to go back through later. Then I took pigment samples from many of the images including greens, blue, red, orange and tan. Then, I made graph comparisons between the pigments of different sections. This will hopefully answer some of the questions from Robert/Bob's email.

First we had a short meeting with Joe. Then Emily and I began going over what we had learned last week. We then read over the email of questions that we wanted to try to answer about the map and began looking at the sections with green correction to try and find what is under them. We ran several different PCA and SAM tests to try and pull out anything interesting, focusing on Korea, Taiwan and Borneo.

First we had our morning meeting with Joe, then we met with our lab groups. We learned about supervised and unsupervised classification and how to run it in ENVI. Then we began working on data from the map while trying to figure out how to use ENVI . We focused on Korea for the majority of the day just trying different analysis tools like SAM, PCA and the classifications, and then moved to Taiwan for the last hour.

First we had a meeting to go over the basic structure of the internship and what to expect. Then we went to the barn to do team building games to get to know everybody. In the lab group we learned the basics of the reflection of light and comparison of colors. We learned the basic explanation of wavelengths and vectors plotted based on the color of light. We looked at the software and analysis that we will be doing but will learn more and start processing tomorrow!