RIT CIS Summer Internship

Joe showed us some of the presentations from previous years during our staff meeting today. We also went down to to the second floor to take a look at the room where our final presentations will be held in. After, my lab partner and I revised some of the slides in our PowerPoint so that the content was more visible to the audience. We added some pictures as well. I continued working on the camera image of sand in ImageJ for awhile. At 10:30, Amy and I went down to the fishbowl to listen to the PhD thesis defense on Ultrafast Laser Polishing for Optical Fabrication. The PhD student did a good job explaining a lot of the jargon. In the afternoon, my lab partner and I exported the data from ImageJ of the images we completed so far into Excel on a different laptop and converted the units from pixels to millimeters and then to microns (since our sieve sizes were in microns originally). With Chris's help, we were able to sort the long lists of particle diameters from smallest to largest

During the staff meeting today, Joe read through my group's presentation outline and gave Hannah and I some feedback. He also suggested that we decrease the word count on some slides in our presentation. In the morning, I revised some of the slides in our PowerPoint and finished analyzing both the microscopic image of clay & silt (image attached) as well as an image of gravel taken using a camera on ImageJ. I also entered the data from the sieving I did yesterday into Excel to generate a graph of sieve size vs. percentage weight. At 11, Amy and I went to a MS thesis defense on shadow detection in aerial imaging over at the Gleason building. The presenter talked about various methods of detecting shadows in an image, including both neural and convolutional neural networks (CNN's), and compared the results of all the methods mentioned. After lunch, my lab partner and I ran two more trials of dry sieving with the gravel (We are officially done with dry sieving!). We generated

This morning, Joe went over everyone's outlines, except for my group's, because he couldn't open the document. Since my lab partner wasn't here today, I ran a second trial of the gravel sample with larger sieve sizes by myself and recorded the weights in a table I created with the new sieve sizes. I took pictures of the gravel using the camera. I also continued to work on the microscopic image of clay & silt in ImageJ. After lunch, I added the images we've taken so far of the samples using both the camera and microscope to our PowerPoint and refined the background slide for automated image analysis. I worked on the image analysis some more, I am very close to finishing that image. Today was pretty successful overall.

During our staff meeting this morning, Joe had one of the students at the Center for Imaging Science come in and present his senior project to us. He talked about OCR's (Optical Character Recognition). After I got to my lab, I worked on the final PowerPoint presentation. I finished the future work slides. I also continued working on the manual image analysis on ImageJ of the clay & silt that I started yesterday. I got a good portion done. Before I left, my lab partner and I picked out new sieve sizes for trials 2 and 3 of the gravel that I will be conducting on Monday. We finished almost all of the dry sieving this week and will be focusing on the manual image analysis and data interpretation next week to add to our presentation.

Today, my lab partner and I sieved a 400 gram sample of gravel. We found that almost all of the gravel was retained on the very top sieve after putting the sample through the sieve shaker. Therefore, we didn't do any more trials with these sieve sizes. We are considering using larger sieve sizes on the sample for trials 2 and 3. After the sieving, I started the manual image analysis using ImageJ for one of the microscopic images of the clay&silt (image attached below). I will continue working on it when I get home as well.

After our staff meeting today, we all went to David's lab to learn more about what he was doing and the goal of his project. He talked about convolutional neural networks (CNN's) and open set recognition. Following this, Joe talked to Hannah and I about our outline and gave us some valuable feedback. Once we got to our lab, we got started right away with bringing materials up to the roof lab for sieving. We completed two trials of the dry sieving for the silt & clay mixture and collected data. We will be experimenting with gravel tomorrow.

In the morning, all the interns went downstairs to Amy's lab. She explained what she worked on last week, which was essentially using python coding to categorize various images of leaves into 3 categories; 1 was healthy and the other 2 were different diseases. Later, my lab partner and I sieved a 400 gram sample of a mixture of clay and silt. This was our first trial for the clay and silt; we need to perform 3 trials for each type of soil for the dry sieving. After the sieving, I took some pictures of the clay and silt sample using the microscope with a mini ruler in the image for manual image analysis purposes later on. I worked on our final PowerPoint presentation as well, adding a few slides. We have all of our background slides completed and some of the procedure and data section done as well.

We didn’t have a staff meeting in the morning today so I got to RIT around 9. My lab partner and I wrote up a draft of an outline for our final PowerPoint presentation and had our advisor check it over. He added some minor details. We also ran the third and final trial of dry sieving for a 400 g sand sample. Hannah weighed the samples/sieves while I recorded the measurements in a data chart. We plan on finishing the dry sieving experiments with silt&clay and gravel samples by the end of this week or the beginning of next week as well as starting the manual and automated image analysis of the different grains using ImageJ and Python coding. I left early today because I have a college essay class in the afternoon everyday this week.

This morning, after a brief staff meeting, all the interns went down to Akul's lab and learned about the EPR machine. He showed us the different paints that they tested and explained how the machine operated. After we got to our lab, Hannah and I performed another test of a 400 gram sample of sand using the sieve shaker with all the variables held constant from yesterday. Amy came and helped out because her advisor wasn't here yet and he didn't leave any work for her to do. I weighed the sieves and sand samples before and after the sieving while Hannah recorded the measurements. At noon, I went to a seminar by Professor Roger Dube on space weather storms (image below). He talked about CME's (coronal mass ejections) and also mentioned the Carrington event, which was a huge solar storm during 1859. I also learned that it takes 10 years to fully recover from the damage done by a space weather storm. In the afternoon, my lab partner and I took pictures of the sand sample th

In the morning, Hannah and I showed the other interns our lab and I explained what the GRIT machine is used for and its individual parts with Chris's help. After, I took another look at the coding left from yesterday. However, since Chris wasn't in the lab at the time we decided to run a trial of my dry sieving with a 200 gram sample of sand from the cabinet first. We took pictures of the sample before and after putting it through the sieve shaker for image analysis purposes later next week. We also made data charts from the different weights of the sieves with and without samples in them. While we were taking apart the sieve stack after shaking the sample, we noticed that there was no sample left in most of the larger sieves so we decided to experiment with a larger (400 gram) sample of the same type of sand after lunch. In the afternoon, Chris also showed me a website that I could refer to to help me with thresholding a grayscale image (a grayscale image and its thresholded c

This morning, Varun showed us what he was doing for his project, which was essentially using Inventor to create 3D models of cars, buildings, and trees for his advisor to use in research. DIRSIG was another application that they used, which simulates what a sensor would detect given the atmospheric, environmental, and optical conditions that are inputted. After Hannah and I got back to our lab, Chris gave a detailed explanation of the GRIT (Goniometer at RIT), including all of its parts and the applications of the data collected. The GRIT machine basically measures the bidirectional reflectance properties of materials. He then turned on the GRIT machine and initiated a scan of a soil sample. It was interesting to watch the graphs on the computer change as the camera moved around the sample to capture reflectance. (Below is an image of the GRIT) In the afternoon, Hannah and I worked on creating a python code for detecting blobs within an image of a soil sample. Chris helped us when our

In the morning, Joe read over Amy's abstract and gave her some feedback. Then, we all went down to Brian's lab on the second floor. He explained his project to us and showed us the eye tracking device that he uses, which tracks the movement of your pupils. The research done in the visual perception lab with this device could be used to track how people behave in the presence of money and where their eyes looked. There are specific symbols on paper money that distinguish real currency from their fake counterparts. This device can track where experts look specifically for the symbols. After Brian showed us around his lab, we all went to our labs. Chris looked over my finalized test plan and made some minor changes. Before lunch, since we had nothing else to do, Hannah and I decided to do a test run of the sieve shaker with a sand sample we found in one of the cupboards. We used the entire container of sediment, which weighed about 997 grams. We turned on the sieve shaker for a li

In the morning, Joe read through all of our abstracts, except for Amy's because she wasn't here, and gave us feedback. Joe also told us that every day this week, we would visit a different lab to get an idea of what our peers were researching. After the meeting, Hannah and I went to our lab and started working on our test plans that our advisor suggested we write-up before starting our experiments. I finished the materials section and most of the procedure section before lunch. I read a few articles to help me with the details of the procedure. Hannah and I also figured out what sieve sizes and soil types we would use for our wet and dry sieving experiments using the sieve shaker. We decided to conduct the same experiments three times to ensure the accuracy of our results and to better compare our data. After lunch, I had Chris look over my test plan outline that I had so far. Later, he sent me a video that described how to use ImageJ to find the individual particle sizes withi

Problem Statement/Abstract The overall goal of my project is to employ physical measurement methods and image analysis techniques on soil samples to validate known measurements such as density and grain size. Ultimately, the results obtained by this research could be used to measure the trafficability, stability, and compaction of a surface. There are many known methods to measure particle size distribution (PSD) within a sample, including wet and dry sieving, laser diffraction, X-ray attenuation, manual and automated image analysis techniques, etc... In this research project, we will be focusing on the different sieving methods in addition to both manual and automated image analysis. More specifically, I will be comparing the dry sieving method with imaging techniques (manual and automated). Dry sieving is carried out using an analytical sieve shaker and measures the PSD based on the percentages of the weight retained within the different sieves. Dry sieving segregates grains by t

This morning, Joe checked in with each of us and asked about our progress in the labs. When we got to our lab, Hannah and I helped our advisor measure some of the plastic bags that needed to be ordered and look on amazon for materials/equipment that were needed for our experiments. At noon, all of the interns attended a seminar on hyper spectral imaging by David Messinger, the head of the imaging science department. He talked about the analysis of the Gough Map (shown below) of Britain using hyper spectral imaging techniques. I found the talk and the powerpoint slides to be very interesting. After, our advisor, Chris, took a look at our abstracts and gave Hannah and I some feedback. We didn't have to change much of what we had originally written which was good! Later in the afternoon before I left, Hannah and I helped Chris with checking that all the of the lab equipment got back from Australia (where it was previously for a couple months). We opened numerous pelican cases to check

During the routine meeting this morning, Joe showed us two videos that featured RIT undergrad students in the Center for Imaging Science and their freshman projects. It was nice to hear what the students had to say about their college experience and their imaging science class in particular. After, Hannah and I went to go listen to a PhD defense seminar that our advisor told us about across the hall. The speaker was Sanghui Han and her presentation was titled "Spectral Image Utility Analysis Towards Optimization of System Design and Operation." She mainly talked about the comprehensive analysis of spectral image utility. It was very interesting to me that she drew inspiration for her research project from her experiences as a commander in the army earlier in life. Following the presentation, I went back to my lab (the GRIT lab) and read another article that I found online comparing dry sieving and laser diffraction (LD) techniques. In the paper, fine and medium-grained marine

As usual we started the day off with a meeting at 8:45. Joe asked each of us how our research was going so far and what our topics were. After, we went to our labs. I finished reading two of the articles that I printed yesterday. The first article was titled "particle size distribution analysis of coarse aggregate using digital image processing"and described in detail both the DIP technique and the mechanical sieving method as well as why the DIP method is better than mechanical sieving. The second article mainly talked about the the effects of different sieving methods (wet and dry) on carbon and nitrogen pools in soils. The study described in the article concluded that while dry sieving resulted in a greater proportion of C associated with fine (less than 150 microns) aggregate (grain) size classes, wet sieving showed greater ability in differentiating soil C associated with coarse aggregate (grain) size classes. Around noon, Hannah, Amy, and I attended a seminar of one of

Today was pretty successful overall. Joe had a brief meeting with all of the interns in the morning and then set us loose to go meet with our advisors and start our research. My advisor gave Hannah and I a quick presentation of the background information associated with what we would be doing over the summer, which includes comparing wet and dry sieving methods as well as both manual and automated image analysis. After, we were given three articles to read that pertained to the experimentation we would be doing in the near future as well as our final project. I was able to finish reading two of the articles and start reading the last one. The first article described in detail the wet and dry sieving methods and highlighted the important features of both. The second article was about automated image-processing techniques that contributed to a better understanding of grain-size distribution within a soil sample and why automated image analysis is better than its manual counterpart. I def

Today was the first day of my internship. We had a brief meeting in the morning and then went to get our ID's, computer accounts, and parking permits. After, Joe explained a team-building activity that we would be working on for the rest of the day: a scavenger hunt. All of us were given the same packet of objects/people that we needed to find around campus. In the end, we were tasked with making a movie that consisted of all the video footage we took of the different objects/people. At 4pm, Joe and two other judges evaluated our project and calculated the total points we received as a group. The scavenger hunt was a good way for us to get to know each other and discover more about the campus. Finally, before we went home, we each met with our advisors and learned more about what we would be doing throughout the course of the internship.