Lab Software Development 2 - free-carrier DIFFUSION analysis system

Quick screenshot:

Paper using the software: https://www.nature.com/articles/s41467-021-25311-1


Motivation:

This is a much more challenging project than the free-carrier lifetime analysis system even though this one does not include hardware control or communication. The free carrier diffusion measurement was already established and is running well. However, it is always very troublesome to do the diffusion length extractions, as our tools require three major fittings - signal response curve fitting, laser profile fitting, and convoluted Gaussian diffusion fitting. It is not possible to find a convenient commercial tool in the market to do the analysis, and it was very time-consuming to do manual analysis with Origin, Python, and other tools. A GUI would be much more convenient to use and can without a double boost the working efficiency of the lab.


Environment:

Since there is no control system needed, there is no need to use the lab-oriented tool LabView. Instead, I chose QtCreator and Python as parts of the analysis system algorithms was achieved with Python already, which was already a lot of programming work.


Functionality - Fit response curve

Our measurement tool (Microwave Impedance Microscopy, or MIM) is quite unique. It yields two channels of signals while both contain parts of the information of local conductivity. To really extract the needed data, we need to use a response curve for conversion. Fitting response curve involves a phase diagram as shown below. If the data points are spreading around the response phase curve, we claim that the fitting meets the requirement. Since users usually measure multiple diffusion maps under different laser intensities, we use the file structure with “Sub-directories” to make everyone’s life easy. Basic I/O schemes are implemented in loading and saving files.

Functionality - Laser profile / Diffusion fitting

Fitting laser or diffusion profile with gaussian or convoluted gaussian functions and extracting laser width or diffusion length is extremely tricky practically in two ways. First, the profiles, especially diffusion profiles are usually not perfectly symmetric. Taking one line across the center may keep a lot of noise in the analysis. It would be ideal if users can take the average with multiple lines radially selected. The second tricky thing is that the center of the profile is not necessarily always exactly at the center. It would be ideal if users can freely select profile centers for different mappings. Solving the two tricky issues required interactive GUI programming skills and took me some time to figure everything out. Below is an example of the analysis. By clicking the center of the laser profile and choose 5 averaging lines, the software evenly chooses target lines and transfer the averaged data into the fitting window. By fine tuning the parameters of the fitting curve, we can easily, and repeatedly, do quick fitting as we want. Fine tuning the center point is now also possible to achieve a better and more solid analysis.

Functionality - Undo Piezo-distortion

This functionality is designed to solve one issue: distortion of the piezo stage which controls x and y scanning range. Basically, the piezo stage may not respond to the voltage in a perfectly linear way, which will cause distortion of the graph if we plot it with voltages as coordinates. In order to get rid of the piezo distortion, we need to do some calibration measurements with standard samples with known distances between structures. It’s obvious that the algorithm needed for removing distortion should be at least in the second order. With the fine tuned parameters while observing the distances of the dots as shown in the figure below, we can get the wanted parameters for the un-distortion process for the real data. Then we can convert the real data into the undistorted one as two figures below. After the user gets the coefficients, automatic conversion of all needed data can be done in the last sector of the tab. This conversion itself can intuitively take as long as O(n^4) time complexity. With indices as locators, I managed to reduce the time complexity to O(n^3). (see resample() method in utils.py)

Summary

Up until now, all basic functionalities needed for quick analysis are finished. There can be more add-ons for the software such as the "map file conversion" which converts between tiff and txt formats, and a separated "Diffusion Length Fitting" with more specific parameters to tune. Overall, this project, even without controlling hardware, is much time-consuming and requires more solid algorithm improvements. With the software done, the users including me myself definitely benefit a lot and can finally get away from the redundant data processing schemes with other primitive data analysis tools.


This project is open source. Feel free to use or improve it if needed.

Github link: https://github.com/xuejianma/LabTools/tree/master/DiffusionAnalysis


Comments

Post a Comment

Popular posts from this blog

Digit sequence detection of SVHN dataset

Image
  Unlike single digit detection problem like MNIST, images in SVHN dataset have more than one digit in series. SVHN Samples. Ref: http://ufldl.stanford.edu/housenumbers/ In this project, we used MSER detection and modified pre-trained VGG16 convolutional neural network to predict the digit sequence given a random image in SVHN dataset. Besides VGG16, we also used non-pre-trained VGG16 and self-built a much smaller CNN as comparisons. VGG16 Architecture. Ref: https://neurohive.io/en/popular-networks/vgg16/  Self-built CNN  The pipeline of the prediction is as: MSER detect blob regions →  CNN  find ROIs(regions of interest) → Combine overlapped/near ROIs →  CNN  Find Digits For training purposes, we mark each image with five integers as the label. The first four integers represent digits in the image, with 0 as no digit for that position and 1-10 representing 1-9 and 0. The fifth integer represents whether the area has any digit, with 0 represent a non-digit area and 1 as a window with d
Read more

Lab Software Development 3 - Scanning Probe Microscope Controller

Image
Motivation     For home-build scanning probe systems in labs, SpecsGroup's Nanonis system is often considered a good solution with various working modes for piezo stage control. However, its' price may not be very friendly for some SPM systems, which do not need a sophisticated feedback PID loop. It is overkill especially for the non-feedback laser microwave impedance microscopy system in our lab. We had a LabView program developed by former students, but it lacks the rotational degree of freedom and line scan mode. Also, the UI's inconvenience has always been painful for users. In order to add more flexibility, I rebuilt a new control system for the SPM with QtCreator with Python as the backend. Functionalities     For convenience, all functionalities are embedded on the same page. Multithreading was implemented for scanning, imaging, and data saving. Many of the functionalities are following the workflow of Park AFM system.      Scan: By clicking "Scan", the con
Read more

Traffic-Sign Recognition Project: Basis for Self-Driving Vehicle Sensing System

Image
This is a project of cs6476 computer vision I am taking. The project is based on the open-cv library and focuses on stationary scenarios with/without noises. It is only allowed to use geometry and basic Hough methods such as HoughCircles, HoughLines, HoughLinesP to recognize the followings: traffic lights and their status, “red”, “yellow”, or “green”; stop signs; yield signs; no-entry signs; warning signs; construction signs. Details of implementation and code are not allowed to share according to student integrity rules of the course. Here only displayed the results under artificial and real images. Detected Green Traffic Light (Artificial Clear Image) Detected Traffic Signs (Artificial Clear Image) More Detected Traffic Signs (Artificial Clear Image) Detected Traffic Signs (Artificial Noisy Image) Detected Stop Sign (Real Image) Detected Yield Sign (Real Image) Detected Warning and Yield Signs (Real Image Stitched)
Read more