hi_dis_config

Description

Defines the DIS configuration information.

Prototype

typedef struct {
    hi_dis_mode         mode;
    hi_dis_motion_level motion_level;
    hi_dis_pdt_type     pdt_type;
    hi_u32           buf_num;
    hi_u32           crop_ratio;
    hi_u32           frame_rate;
    hi_bool camera_steady;
    hi_bool scale;
} hi_dis_config;

Members

Member	Description
mode	Multi-degree DIS algorithm. There are three algorithms in total.
motion_level	Motion level of the camera.
pdt_type	Product form with the DIS function. The IPC, action camera, and drone camera are supported currently.
buf_num	Number of memory blocks used by the DIS to store images. The number of buffers can be increased if frame loss occurs for the DIS output frames. Value range: [5, 10]
crop_ratio	Indicates the cropping ratio of the DIS output image. Value range: [50, 98]
frame_rate	Set to the actual frame rate of the VI output. Value Range: (0, 60]. When the DIS mode is HI_DIS_MODE_4_DOF_GME or HI_DIS_MODE_6_DOF_GME, a low frame rate increases the interval between two frames and increases the jitter amplitude between two frames. In addition, the rolling shutter effect is intensified. As a result, DIS may be compromised, especially in scenarios with large jitter amplitude. When the DIS mode is HI_DIS_MODE_GYRO, DIS may be compromised as that in the GME mode, and the maximum sampling rate of the gyroscope supported by the algorithm decreases. The current 3.5 kHz is based on 25 FPS (the minimum frame rate). If the value decreases, the maximum sampling rate of the supported gyroscope also decreases.
camera_steady	Whether the camera is steady. The options are as follows: HI_FALSE: not steady HI_TRUE: steady
scale	Whether to scale up the output picture after cropping. The current DIS function allows the user to choose whether to scale up the output picture after cropping. If you choose not to, you can use the backend VPSS for upscaling. The options are as follows: HI_FALSE: not upscaled HI_TRUE: upscaled When scale is set to HI_FALSE, the width and height of the output image are the width and height of the input image multiplied by the cropping ratio configured by crop_ratio, and the value is 2-pixel aligned.

Parent topic: Video Input (VI)