TI Autonomous Driving Algorithms (TIADALG) Library User Guide
tiadalg_dof_plane_seperation.h File Reference

This file defines the interface for creating three plane output from packed output from DOF h/w. More...

#include <stdint.h>

Go to the source code of this file.

Functions

int32_t tiadalg_dof_plane_seperation_cn (uint32_t *dof_plane, int32_t img_width, int32_t img_height, int32_t out_data_type, int32_t pad_pixel[4], void *out_rgb_planes, void *out_yuv_planes)
 It does 3 plane creation from raw Dense Optical Flow (DOF) generated by DMPACK hardware accelerator. This API is intended to do generate 3 planes of Dof to be used in Deep Learning networks. First plane is 8-bit horizontal flow vector, second plane is 8-bit vertical flow vector, and last plane is 8-bit confidence for flow vectors. More...
 
int32_t tiadalg_dof_plane_seperation_c66 (uint32_t *dof_plane, int32_t img_width, int32_t img_height, int32_t out_data_type, int32_t pad_pixel[4], void *out_rgb_planes, void *out_yuv_planes)
 It does 3 plane creation from raw Dense Optical Flow (DOF) generated by DMPACK hardware accelerator. This API is intended to do generate 3 planes of Dof to be used in Deep Learning networks. First plane is 8-bit horizontal flow vector, second plane is 8-bit vertical flow vector, and last plane is 8-bit confidence for flow vectors. More...
 

Detailed Description

This file defines the interface for creating three plane output from packed output from DOF h/w.

Version
0.1 Sep 2018 : Initial Code

Function Documentation

◆ tiadalg_dof_plane_seperation_cn()

int32_t tiadalg_dof_plane_seperation_cn ( uint32_t *  dof_plane,
int32_t  img_width,
int32_t  img_height,
int32_t  out_data_type,
int32_t  pad_pixel[4],
void *  out_rgb_planes,
void *  out_yuv_planes 
)

It does 3 plane creation from raw Dense Optical Flow (DOF) generated by DMPACK hardware accelerator. This API is intended to do generate 3 planes of Dof to be used in Deep Learning networks. First plane is 8-bit horizontal flow vector, second plane is 8-bit vertical flow vector, and last plane is 8-bit confidence for flow vectors.

tiadalg_dof_plane_seperation_cn

Parameters
dof_plane[in] : Input row buffer given out by dof hardware accelerator
img_width[in] : Width of image for which flow vector has been provided
img_height[in] : Height of image for which flow vector is provided
out_data_type[in] : Output plane data type. Currently only U08 is supported
pad_pixel[in] : Padding of the output plane. 4 values for Left, Top, Right, Bottom needs to be provided.
out_rgb_planes[in] : Buffer for output RGB plane.
out_yuv_planes[in] : Buffer for output semi-planer yuv 420 planes.
Note
  • User should allocate output buffer and it should be of size (img_width + pad_pixel[0] + pad_pixel[2])* (img_height + pad_pixel[1] + pad_pixel[3]) for out_data_type of U08.
  • Size of the input buffer should be img_width x img_height

◆ tiadalg_dof_plane_seperation_c66()

int32_t tiadalg_dof_plane_seperation_c66 ( uint32_t *  dof_plane,
int32_t  img_width,
int32_t  img_height,
int32_t  out_data_type,
int32_t  pad_pixel[4],
void *  out_rgb_planes,
void *  out_yuv_planes 
)

It does 3 plane creation from raw Dense Optical Flow (DOF) generated by DMPACK hardware accelerator. This API is intended to do generate 3 planes of Dof to be used in Deep Learning networks. First plane is 8-bit horizontal flow vector, second plane is 8-bit vertical flow vector, and last plane is 8-bit confidence for flow vectors.

tiadalg_dof_plane_seperation_c66

Parameters
dof_plane[in] : Input row buffer given out by dof hardware accelerator
img_width[in] : Width of image for which flow vector has been provided
img_height[in] : Height of image for which flow vector is provided
out_data_type[in] : Output plane data type. Currently only U08 is supported
pad_pixel[in] : Padding of the output plane. 4 values for Left, Top, Right, Bottom needs to be provided.
out_rgb_planes[in] : Buffer for output RGB plane.
out_yuv_planes[in] : Buffer for output semi-planer yuv 420 planes.
Note
  • User should allocate output buffer and it should be of size (img_width + pad_pixel[0] + pad_pixel[2])* (img_height + pad_pixel[1] + pad_pixel[3]) for out_data_type of U08.
  • Size of the input buffer should be img_width x img_height

© Copyright 2018 Texas Instruments Incorporated. All rights reserved.
Document generated by doxygen 1.8.6