Creating a 2D scatter plot with thumbnail images
$ python3 ./thumb_scatter.py --help
thumb_scatter.py - making a scatter plot with thumbnail images
Usage:
thumb_scatter.py [-h|--help] [--scale=scale_value] input_csv_file plot_title output_file
Options:
--scale : size of thumnail (scale value as a real number : default = 1.0)
--emphasize : Class to be emphasize (4 digit hexadecimal code)
--color : colorize character frames according to their colasses
-h or --help : display this help message
The input CSV file contains one sample par one line in the following format.
''' x, y, image file name '''
Please see sample.csv.
2.35,1.89,sample_imgs/a001.png
4.59,-2.08,sample_imgs/a002.png
-3.2,-0.0116,sample_imgs/a003.png
-1.24,3.64,sample_imgs/a004.png
-1.02,-0.0455,sample_imgs/a005.png
python3 ./thumb_scatter.py --scale=2.5 sample.csv --color 'sample scatter plot' sample.pdf
If sample number in the CSV is too may to see the result. You can reduce the number by the following command.
awk 'NR%2==0{print $0}' sample.csv > re_sample.csv
The command output only even samples, so the number of samples is reduced as 1/2.
Creating ranking list image
$ python3 ./rank_images.py Input_CSV_File
The input CSV file contains multiple data blocks. One data block consits of multiple lines.
<filename of query image>
<filename of rank-n image>,<distance value> (repeated n-times)
"
" (delimiter of image block)
Example:
resize/0812_0614086.png
resize/0813_0614087.png,26.48101234436035
resize/1108_0622215.png,29.828632354736328
resize/1748_0625685.png,30.70377540588379
resize/0635_0610158.png,32.40945053100586
resize/0022_0601027.png,33.12775421142578
"
"
resize/1520_0625400.png
resize/1435_0625299.png,30.205413818359375
resize/0208_0622363.png,32.262229919433594
resize/0240_0604146.png,32.40801239013672
resize/1476_0625349.png,33.261756896972656
resize/0560_0610020.png,33.75419616699219
"
python3 ./rank_images.py ./Img_Dist/cae_L2.csv
NOTE The program outputs result images in the directory where the CSV file is.