Metadata-Version: 2.1
Name: selective_search
Version: 0.1.2
Summary: Selective Search in Python
Home-page: https://github.com/ChenjieXu/selective_search
Author: Chenjie Xu
Author-email: cxuscience@gmail.com
License: MIT
Description: # Selective Search
        [![GitHub release](https://img.shields.io/github/v/release/ChenjieXu/selective_search?include_prereleases)](https://github.com/ChenjieXu/selective_search/releases/)
        [![PyPI](https://img.shields.io/pypi/v/selective_search)](https://pypi.org/project/selective-search/)
        [![Travis Build Status](https://travis-ci.org/ChenjieXu/selective_search.svg?branch=master)](https://travis-ci.org/ChenjieXu/selective_search)
        [![Codacy grade](https://img.shields.io/codacy/grade/8d5b9ce875004d458bdf570f4d719472)](https://www.codacy.com/manual/ChenjieXu/selective_search)
        
        This is a full implementation of selective search in Python. The implementation is typically based on this paper[[1]](#Uijlings). It have three selective search modes according to various diversification strategies as in the paper.
        
        ## Installation
        Installing from [PyPI](https://pypi.org/project/selective-search/) is recommended :
        ```
        $ pip install selective-search
        ```
        It is also possible to install the latest version from [Github source](https://github.com/ChenjieXu/selective_search/):
        ```
        $ git clone https://github.com/ChenjieXu/selective_search.git
        $ cd selective_search
        $ python setup.py install
        ```
        
        ## Quick Start
        
        ```python
        import skimage.io
        from selective_search import selective_search
        
        # Load image as NumPy array from image files
        image = skimage.io.imread('path/to/image')
        
        # Run selective search using single mode
        boxes = selective_search(image, mode='single', random_sort=False)
        ```
        For detailed examples, refer [this](https://github.com/ChenjieXu/selective_search/tree/master/examples) part of the repository.
        
        ## Parameters
        
        ### Mode
        
        Three modes correspond to various combinations of diversification strategies. The appoach to combine different diversification strategies, say, color spaces, similarity measures, starting regions is listed in the following table[[1]](#Uijlings).
        
        | Mode    | Color Spaces        | Similarity Measures | Starting Regions (k) | Number of Combinations |
        |---------|---------------------|---------------------|----------------------|------------------------|
        | single  | HSV                 | CTSF                | 100                  | 1                      |
        | fast    | HSV, Lab            | CTSF, TSF           | 50, 100              | 8                      |
        | quality | HSV, Lab, rgI, H, I | CTSF, TSF, F, S     | 50, 100, 150, 300    | 80                     |
        
        *  **Color Space** [[Source Code]](https://github.com/ChenjieXu/selective_search/blob/master/selective_search/util.py#L23)  
          Initial oversegmentation algorithm and our subsequent grouping algorithm are performed in this colour space.
        
        *  **Similarity Measure** [[Source Code]](https://github.com/ChenjieXu/selective_search/blob/master/selective_search/measure.py#L101)  
          'CTSF' means the similarity measure is aggregate of color similarity, texture similarity, size similarity, and fill similarity.
          
        *  **Starting Region** [[Source Code]](https://github.com/ChenjieXu/selective_search/blob/master/selective_search/util.py#L9)  
          A parameter of initial grouping algorithm[[2]](#Felzenszwalb), which yields high quality starting locations efficiently. A larger k causes a preference for larger components of initial strating regions.
        
        ### Random Sort
        
        If random_sort set to True, function will carry out pseudo random sorting. It only alters sequences of bounding boxes, instead of locations, which prevents heavily emphasis on large regions as combing proposals from up to 80 different strategies[[1]](#Uijlings). This only has a significant impact when selecting a subset of region proposals with high rankings, as in RCNN.
        
        ## References
        
        \[1\] <a name="Uijlings"> [J. R. R. Uijlings et al., Selective Search for Object Recognition, IJCV, 2013](https://ivi.fnwi.uva.nl/isis/publications/bibtexbrowser.php?key=UijlingsIJCV2013&bib=all.bib)  
        \[2\] <a name="Felzenszwalb"> [Felzenszwalb, P. F. et al., Efficient Graph-based Image Segmentation, IJCV, 2004](https://ivi.fnwi.uva.nl/isis/publications/bibtexbrowser.php?key=UijlingsIJCV2013&bib=all.bib)  
        \[3\] <a name='koen'> [Segmentation as Selective Search for Object Recognition](https://www.koen.me/research/selectivesearch/)
        
Keywords: rcnn
Platform: UNKNOWN
Classifier: Development Status :: 5 - Production/Stable
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Software Development :: Libraries
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Education
Classifier: Intended Audience :: Science/Research
Description-Content-Type: text/markdown
