vpc_get_perspective_lut

0: conventional bilinear algorithm, whose compute process is close to that of OpenCV.
1: conventional nearest neighbor algorithm, whose compute process is close to that of OpenCV.

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product/Atlas A3 inference product	√
Atlas A2 training product/Atlas A2 inference product	√
Atlas training product	x
Atlas inference product	x
Atlas 200I/500 A2 inference product	√

Obtains a pixel position remapping LUT of perspective transformation according to the pixel position coordinates in the input and output images.

C Prototype

hi_s32hi_mpi_vpc_get_perspective_lut(hi_point_pair_info *point_pair_info, hi_u32  interpolation, hi_remap_lut *remap_lut)

Python Function

remap_lut, ret = acl.himpi.vpc_get_perspective_lut(point_pair_info, interpolation, remap_lut)

Parameter	Description
point_pair_info	Dict, contains four pairs of pixel coordinates in the input and output images. Each pair contains pixel coordinates of an input image and an output image, respectively. For details, see hi_point_pair_info.
interpolation	Int. The following resizing algorithms are supported: 0: conventional bilinear algorithm, whose compute process is close to that of OpenCV. 1: conventional nearest neighbor algorithm, whose compute process is close to that of OpenCV.
remap_lut	dict, dictionary of the pixel position remapping LUT for perspective transformation. For details, see hi_remap_lut. You need to allocate buffer for the lut attribute in advance. The lut_size attribute of the buffer size can be obtained by calling acl.himpi.vpc_get_lut_mem_size.

Return Value	Description
remap_lut	Dict, dictionary of information about the pixel position remapping LUT of perspective transformation. After this API is successfully called, you can obtain the LUT information of affine transformation by using the address in the lut attribute.
ret	Int, error code. 0 indicates success. Other values indicate failure. For details, see VPC Return Codes.

Parent topic: VPC APIs