Class GaussianBlur

Defined in File vision_lite.h

Inheritance Relationships

Base Type

public mindspore::dataset::TensorTransform (Class TensorTransform)

Class Documentation

class GaussianBlur : public mindspore::dataset::TensorTransform 

Blur the input image with the specified Gaussian kernel.

Public Functions

explicit GaussianBlur(const std::vector<int32_t> &kernel_size, const std::vector<float> &sigma = {0., 0.})

Constructor.

Example

/* Define operations */
auto decode_op = vision::Decode();
auto gaussian_op = vision::GaussianBlur({7, 7});

/* dataset is an instance of Dataset object */
dataset = dataset->Map({decode_op, gaussian_op},  // operations
                       {"image"});                // input columns

参数

kernel_size – [in] A vector of Gaussian kernel size for width and height. The value must be positive and odd.
sigma – [in] A vector of Gaussian kernel standard deviation sigma for width and height. The values must be positive. Using default value 0 means to calculate the sigma according to the kernel size.

~GaussianBlur() = default: Destructor.