1. How to reconstruct the three-dimensional shape of an object in a scene in a very inexpensive way using shadow-casting
2. How to write and execute a matlab script from a bash script
I've written a bash script - shadowscan.sh - which is a wrapper for a set of programs for the reconstruction 3D surface of an object in a series of photographs using the ingenious shadow-casting algorithm of Jean-Yves Bouguet and Pietro Perona
The script relies heavily on the GUI-style interfacing tools of Zenity and acts as an interface and wrapper for the algorithm as implemented by these guys. It will also write and execute a matlab script for visualising the outputs.
This script should live in a directory with the following compiled programs, all available here:
1. ccalib (performs the camera calibration to retrieve the intrinsic parameters of camera and desk)
2. find_light (performs the light calibration to find light-source coordinates)
3. desk_scan (does the 3d reconstruction)
4. merge (merges 2 3d reconstructions, e.g. from different angles)
and the following matlab script:
1. selectdata.m (available here)
When the makefile is run, it creates the compiled programs in the ./bin directory. It's important to consult the README in the main folder as well as within the 'desk_scan' folder to have a clear idea of what the program does and what the input parameters mean.
My program will check if the camera calibration and light calibration files exist in the folder and if not will carry them out accordingly then several reconstructions are carried out, and the outputs merged. Finally matlab is called from the command line to run 'p_lightscan_data.m' to visualize the results.
when the script finishes there will be the following output files:
1. xdim.txt - max dimension to scan to
2. params - camera calibration parameters
3. data - camera calibration data
4. light.txt - light calibration parameters
5. sample.pts - [x,y,z] of object, in coordinates relative to light source
6. sample_lightscan1.tif - filled colour contour plot of reconstructed object
7. sample_lightscan2.tif - mesh plot of reconstructed object
I've tried to keep it flexible. It will run only the elements it needs to. For example, if the camera or lighting has been carried out before for a different object it will pick up on that. Also, the program can be invoked by passing it two arguments (path to folder where sample images are, and what sequence of contrasts to use for the reconstruction), which avoids having to input with the graphical menus. Example usage:
1. bash shadowscan.sh
2. bash shadowscan.sh /home/daniel/Desktop/shadow_scanning/sample/bottle 30.35.40.45
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| defined() | |
| { | |
| [ ${!1-X} == ${!1-Y} ] | |
| } | |
| args=("$@") # get arguments passed to the program | |
| # if no arguments, IMAGES and CNTRST will be empty otherwise will be filled with the arguments | |
| defined args && IMAGES=${args[0]} # first = sample image location e.g. /home/daniel/Desktop/shadow_scanning/sample/ | |
| defined args && CNTRST=${args[1]} # second = contrst values - no more than four! e.g. 30.35.40.45 | |
| # if IMAGES empty, display flash info, and prompt user to direct program to folder | |
| if [ -z $IMAGES ] ; then | |
| zenity --info --title="SHADOWSCAN" --text "This program will create a 3D reconstruction of an object using images of the object scanned by a shadow. It is a bash/matlab wrapper program for code created by http://www.cs.columbia.edu/~pblaer/projects/photo/ which implements the shadow-casting algorithm of Jean-Yves Bouguet and Pietro Perona (http://www.vision.caltech.edu/bouguetj/ICCV98/)" | |
| # folder where sample images are | |
| IMAGES=$(zenity --file-selection --directory --title="Select directory where sample images are") | |
| fi | |
| if [ ! -f ./param ] # if param does not exist, do cam calib | |
| then | |
| # folder where checkerboard images are | |
| CAL_CAM=$(zenity --file-selection --title="Select checkerboard image" --text "") | |
| # get x-dimension size (number of pixels), subtract 20, and write to file | |
| convert $CAL_CAM -format "%[fx: min(w,h)-20]" info: > xdim.txt | |
| # input size of square in calibration photo # 23 | |
| SQ_HGHT=$(zenity --title="Camera Calibration" --text "input square size" --entry) | |
| ./ccalib $CAL_CAM $SQ_HGHT | |
| fi | |
| if [ ! -f ./light.txt ] # if light.txt does not exist, do light calib | |
| then | |
| # folder where light calibration images are | |
| CAL_LIGHT=$(zenity --file-selection --directory --title="Select directory where light calibration images are") | |
| # input height of the pencil in the light calibration photos # 76 | |
| OBJ_HGHT=$(zenity --title="Light Calibration" --text "input object height" --entry) | |
| zenity --info --title="Instructions" --text "For each image, click first on the shadow tip, then on the base" | |
| ./find_light param $OBJ_HGHT $CAL_LIGHT/*.pgm > light.txt | |
| fi | |
| # this is the reference at the bottom of the image. This | |
| # should be a row where at no time does the shadow pass | |
| # over an object, it always should project onto the desk plane. | |
| if [ ! -f ./xdim.txt ] ; then # if xdim.txt does not exist | |
| BTM=$(zenity --title="reference value" --text "at the bottom of the image" --entry) | |
| else | |
| BTM=$(head xdim.txt) | |
| fi | |
| # this is the reference at the top of the image. This | |
| # should be a row where at no time does the shadow pass | |
| # over an object, it always should project onto the desk plane. | |
| TOP=20 | |
| # if CNTRST empty, prompt user to select which CNTRST to use | |
| if [ -z $CNTRST ] ; then | |
| # user selects contrast levels (1 to 4) to be processed from a limited list | |
| CNTRST=$(zenity --list --text "Select one or more contrast values to process" --title="" --checklist --column "Pick" --column "options" TRUE "30" TRUE "35" FALSE "40" FALSE "45" --separator=".") | |
| fi | |
| # $Var are contrast values, 0 is desk plane | |
| if [ ${#CNTRST} -eq 2 ]; then # 1 contrast | |
| Var1=`echo $CNTRST | awk -F\. '{print $1}'` | |
| ./desk_scan light.txt param $Var1 0 $TOP $BTM sample.pts $IMAGES/*.pgm | |
| else | |
| if [ ${#CNTRST} -eq 5 ] ; then # 2 contrasts | |
| Var1=`echo $CNTRST | awk -F\. '{print $1}'` | |
| Var2=`echo $CNTRST | awk -F\. '{print $2}'` | |
| ./desk_scan light.txt param $Var1 0 $TOP $BTM sample1.pts $IMAGES/*.pgm | |
| ./desk_scan light.txt param $Var2 0 $TOP $BTM sample2.pts $IMAGES/*.pgm | |
| ./merge sample1.pts sample2.pts sample.pts | |
| rm sample1.pts sample2.pts | |
| else | |
| if [ ${#CNTRST} -eq 8 ] ; then # 3 contrasts | |
| Var1=`echo $CNTRST | awk -F\. '{print $1}'` | |
| Var2=`echo $CNTRST | awk -F\. '{print $2}'` | |
| Var3=`echo $CNTRST | awk -F\. '{print $3}'` | |
| ./desk_scan light.txt param $Var1 0 $TOP $BTM sample1.pts $IMAGES/*.pgm | |
| ./desk_scan light.txt param $Var2 0 $TOP $BTM sample2.pts $IMAGES/*.pgm | |
| ./desk_scan light.txt param $Var3 0 $TOP $BTM sample3.pts $IMAGES/*.pgm | |
| # merge the 3 data sets | |
| ./merge sample1.pts sample2.pts sample12.pts | |
| ./merge sample1.pts sample3.pts sample13.pts | |
| ./merge sample2.pts sample3.pts sample23.pts | |
| ./merge sample12.pts sample13.pts sample12_13.pts | |
| ./merge sample12_13.pts sample23.pts sample.pts | |
| rm sample1.pts sample2.pts sample3.pts sample12.pts sample13.pts sample23.pts sample12_13.pts | |
| else | |
| if [ ${#CNTRST} -eq 11 ] ; then # 4 contrasts | |
| Var1=`echo $CNTRST | awk -F\. '{print $1}'` | |
| Var2=`echo $CNTRST | awk -F\. '{print $2}'` | |
| Var3=`echo $CNTRST | awk -F\. '{print $3}'` | |
| Var4=`echo $CNTRST | awk -F\. '{print $4}'` | |
| ./desk_scan light.txt param $Var1 0 $TOP $BTM sample1.pts $IMAGES/*.pgm | |
| ./desk_scan light.txt param $Var2 0 $TOP $BTM sample2.pts $IMAGES/*.pgm | |
| ./desk_scan light.txt param $Var3 0 $TOP $BTM sample3.pts $IMAGES/*.pgm | |
| ./desk_scan light.txt param $Var4 0 $TOP $BTM sample4.pts $IMAGES/*.pgm | |
| # merge the 4 data sets | |
| ./merge sample1.pts sample2.pts sample12.pts | |
| ./merge sample3.pts sample4.pts sample34.pts | |
| ./merge sample12.pts sample34.pts sample.pts | |
| rm sample1.pts sample2.pts sample3.pts sample4.pts sample12.pts sample34.pts | |
| fi | |
| fi | |
| fi | |
| fi | |
| # once completed, in matlab run 'p_lightscan_data.m' to visualize the results | |
| # http://ubuntuforums.org/showthread.php?t=825042 | |
| filename=p_lightscan_data.m; | |
| cat > $filename << EOF | |
| % p_lightscan_data | |
| % example matlab code to clean up and plot data generated from shadowscan.sh | |
| % Daniel Buscombe March 2011 | |
| close all;clear all;clc | |
| pts=importdata('sample.pts'); | |
| plot(pts.data(:,1),pts.data(:,2),'k.') | |
| view(2) | |
| [pl,xs,ys] = selectdata('sel','rect'); | |
| close | |
| x=pts.data(pl,1); y=pts.data(pl,2); z=pts.data(pl,3); | |
| f=find(z==0); x(f)=[]; y(f)=[]; z(f)=[]; | |
| [xi,yi]=meshgrid([min(x):.1:max(x)],[min(y):.1:max(y)]); | |
| zi=griddata(x,y,z,xi,yi); | |
| figure | |
| contourf(xi,yi,-zi-min(-zi(:))), colorbar | |
| print -dtiff -r300 sample_lightscan1.tif | |
| close | |
| figure | |
| mesh(xi,yi,-zi-min(-zi(:))) | |
| print -dtiff -r300 sample_lightscan2.tif | |
| close | |
| EOF | |
| chmod +x $filename | |
| /opt/matlab/bin/matlab -nodesktop -nosplash -r "p_lightscan_data;quit;" |
No comments:
Post a Comment