Metadata-Version: 2.1
Name: yyimg
Version: 1.0.2
Summary: image tools for deep learning
Home-page: https://github.com/YunYang1994/yyimg.git
Author: YunYang1994
Author-email: dreameryangyun@sjtu.edu.cn
License: UNKNOWN
Description: <p align="center">
            <img width="65%" src="https://user-images.githubusercontent.com/30433053/68877902-c5bd2f00-0741-11ea-8cac-af227a77bb14.png" style="max-width:65%;">
            </a>
        </p>
        
        # Introduction
        yyimg is a high-level image-processing tool, written in Python and using [OpenCV](https://github.com/opencv/opencv) as backbend. This repo helps you with processing images for your deep learning projects. 
        
        # Installation
        Commands to install from pip or download the source code from our website https://pypi.org/project/yyimg
        
        ```bashrc
        $ pip3 install yyimg==1.0.0rc
        ```
        
        # Example Useage
        
        Take one image in Kitti dataset for example:
        
        ```python
        import yyimg
        from PIL import Image
        image, boxes, classes = yyimg.load_data()
        ```
        |Items|Description|
        |---|---
        |image|a numpy array of shape (height, width, #channels)
        |boxes|a numpy array of shape (N, 5), representing N 2Dboxes of `[class_index, xmin, ymin, xmax, ymax]`
        |classes|a list of class names
        
        ```python
        print(classes)
        ['Car', 'Truck', 'Van', 'Pedestrian']
        ```
        
        ## visualize 2D boxes
        
        ```python
        draw_image = yyimg.draw_2Dbox(image, boxes, class_category=classes)
        draw_image = cv2.cvtColor(draw_image, cv2.COLOR_BGR2RGB) # BGR -> RGB
        Image.fromarray(draw_image).show()
        ```
        ![image](https://user-images.githubusercontent.com/30433053/69005372-49526800-095c-11ea-8984-4d03154eab80.jpg)
        
        ## data augmentation
        
        ### - horizontal_flip
        
        with 2D bounding boxes:
        ```python
        aug_image, boxes = yyimg.horizontal_flip(image, boxes)
        ```
        without 2D bounding boxes:
        ```python
        aug_image = yyimg.horizontal_flip(image)
        ```
        ![image](https://user-images.githubusercontent.com/30433053/69005858-b668fc00-0962-11ea-89a9-2e06bf14fb2d.jpg)
        
        ### - add_rain
        
        ```python
        aug_image = yyimg.add_rain(image)
        ```
        ![image](https://user-images.githubusercontent.com/30433053/69005561-d0084480-095e-11ea-9b8d-c94f7694585b.jpg)
        
        ### - shift_gama
        
        ```python
        aug_image = yyimg.shift_gamma(image) 
        ```
        ![image](https://user-images.githubusercontent.com/30433053/69005465-c7633e80-095d-11ea-856c-9bc22b213e5c.jpg)
        
        ### - shift_brightness
        
        ```python
        aug_image = yyimg.shift_brightness(image)
        ```
        ![image](https://user-images.githubusercontent.com/30433053/69005494-435d8680-095e-11ea-9922-1ee73571b645.jpg)
        
        ### - shift_color
        
        ```python
        aug_image = yyimg.shift_color(image)
        ```
        ![image](https://user-images.githubusercontent.com/30433053/69005754-6f2e3b80-0961-11ea-9095-ed5c0497dcdc.jpg)
        
        
        
Keywords: deep-learning,image
Platform: UNKNOWN
Description-Content-Type: text/markdown
