DvppDataInfo

Function

Defines the DVPP entity data.

Structure Definition

struct DvppDataInfo {
    uint32_t width = 0;                                    
    uint32_t height = 0;                                  
    uint32_t widthStride = 0;                              
    uint32_t heightStride = 0;                             
    MxbasePixelFormat format = MXBASE_PIXEL_FORMAT_YUV_SEMIPLANAR_420;  
    uint32_t frameId = 0;                                  
    uint32_t channelId = 0;                                
    uint32_t dataSize = 0;
    uint32_t outDataSize = 0;
    uint32_t dataType = 0;                                
    uint8_t* data = nullptr;                              
    uint8_t* outData = nullptr;
    uint32_t resizeWidth = 0;
    uint32_t resizeHeight = 0;
    std::string device = "host:0";                         
    void (*destory)(void *) = nullptr;        
};

Parameter Description

Parameter	Input/Output	Description
width	Input	Width of the source image
height	Input	Height of the source image
widthStride	Input/Output	Width of the source image after alignment
heightStride	Input/Output	Height of the source image after alignment
format	Input/Output	Image format. The default value is MXBASE_PIXEL_FORMAT_YUV_SEMIPLANAR_420. For details about the definition of the DVPP image format, see MxbasePixelFormat.
frameId	Input	ID of an image frame
channelId	Input	ID of an image channel
dataSize	Input/Output	Image data size, in bytes. The size must be the same as the actual image size. Otherwise, a core dump may occur.
outDataSize	Input/Output	Size of the output image data, in bytes. The value of outDataSize must be the same as that of the output image. Otherwise, a core dump may occur.
dataType	Input	Image data type
data	Input/Output	Image data
outData	Input/Output	Address of the pre-allocated memory, which is used to store the image data after video decoding.
resizeWidth	Input	Resizes the width during video decoding. Currently, only Atlas inference products support this parameter. The default value is 0, indicating that no resizing is performed. Value range: [10, 4096]
resizeHeight	Input	Resizes the height during video decoding. Currently, only Atlas inference products support this parameter. The default value is 0, indicating that no resizing is performed. Value range: [6, 4096]
device	Input	Device ID
destroy	Input/Output	Callback function for destroying the DVPP data

Parent topic: Data Structure