一个针对JPEG格式图像提取原始数据，方便图像数据操作的python库
- Mac OS X
- python3
- libjpeg 9c
Mac OS X:
brew install python
brew install libjpeg
>>> python setup.py install
>>> from pyjpegtbx import JPEGImage
>>>
>>> # False for reading RGB data
>>> img = JPEGImage('lfs.jpg', False)
>>> newdata = []
>>> # RGB data now is in img.data
>>> for rgb in img.data:
>>> newdata.append((0, rgb[1], rgb[2]))
>>> img.data = newdata
>>> img.save('tmp0.jpg')
>>>
>>> # True(default) for reading DCT data
>>> img = JPEGImage('lfs.jpg')
>>> # components' info
>>> print(img.comp_infos)
>>> # quality tables
>>> print(img.quant_tbls)
>>> # huffman tables
>>> print(img.ac_huff_tables)
>>> print(img.dc_huff_tables)
>>> print(img.filename)
>>> print(img.size) # (width, height)
>>> # DCT data now is in img.data
>>> # print(img.data)
>>> # set some of DCT data to zero
>>> for key, val in img.data.items():
>>> for coef in val:
>>> for i in range(32, 64):
>>> coef[i] = 0
>>> img.save('tmp1.jpg')
>>> from pyjpegtbx import JPEGImage
>>>
>>> # 第二个参数为False, 表示读取RGB数据
>>> img = JPEGImage('lfs.jpg', False)
>>> newdata = []
>>> # img.data中存储着RGB数据
>>> for rgb in img.data:
>>> newdata.append((0, rgb[1], rgb[2]))
>>> img.data = newdata
>>> img.save('tmp0.jpg')
>>>
>>> # 第二个参数默认为True, 表示读取DCT数据
>>> img = JPEGImage('lfs.jpg')
>>> # 查看颜色分量信息
>>> print(img.comp_infos)
>>> # 量化表
>>> print(img.quant_tbls)
>>> # 哈夫曼表
>>> print(img.ac_huff_tables)
>>> print(img.dc_huff_tables)
>>> # 文件名
>>> print(img.filename
>>> # (width, height)
>>> print(img.size)
>>> # DCT数据信息
>>> # print(img.data)
>>> # 把高频DCT系数设置为0
>>> for key, val in img.data.items():
>>> for coef in val:
>>> for i in range(32, 64):
>>> coef[i] = 0
>>> img.save('tmp1.jpg')
- Thanks for Independent JPEG Group (ijg) providing the library for JPEG image compression. link
- Thanks for these article 使用C语言扩展Python(一)~(五)