whdcumt's repositories
MSR_original
Retinex-IPOL
awesome-deep-learning
A curated list of awesome Deep Learning tutorials, projects and communities.
DeepLearningBook
MIT Deep Learning Book in PDF format
Histogram-Equalization-Logarithmic-Mapping-Image-Rotation-Gaussian-Averaging-Filter-Median-Filter
Implementation of Histogram Equalization, Logarithmic Mapping, Image Rotation, Gaussian Averaging Filter and Median Filter. Histogram equalization is a method using which we can adjust image intensities that will enhance the contrast of the image. By applying the histogram, the intensities of the image can be distributed properly which will allow the area with lower contrast to gain proper higher contrast. To compress the dynamic range of an image we can replace each and every pixel value with their log values. By implementing this, we enhance the low intensity pixels of the image. Logarithmic mapping is a simple point process in which the mapping function is a log curve. Since the logarithm is not defined for, the equation of the log mapping function looks like given below: 𝐼′(𝑟,𝑐)=log(𝐼(𝑟,𝑐)+1). Rotation is the operation that will perform the geometric transformation of the image element which will be rotated at the given angle about an origin. Gaussian filter have a special type of property of not having any overshoot towards a step input function and it minimizes the fall and rise time, thus it has minimum delay in group. Due to this property the Gaussian filter is considered the ideal time domain filter. Median filter is basically used to remove the nonlinear noise in the images. Generally this is used in the preprocessing step in different image processing algorithms. The most significant reason to use the median filtering is that it preserves the edge information even after removing the nonlinear noise.
ImProCv
第一程序:读取图片,灰度化,边缘检测
labelImg
:metal: LabelImg is a graphical image annotation tool and label object bounding boxes in images
PureGaussian
实现模糊功能