Video scaler. More...
Typedefs | |
typedef struct gavl_video_scaler_s | gavl_video_scaler_t |
Opaque scaler structure. | |
Functions | |
GAVL_PUBLIC gavl_video_scaler_t * | gavl_video_scaler_create () |
Create a video scaler. | |
GAVL_PUBLIC void | gavl_video_scaler_destroy (gavl_video_scaler_t *scaler) |
Destroy a video scaler. | |
GAVL_PUBLIC gavl_video_options_t * | gavl_video_scaler_get_options (gavl_video_scaler_t *scaler) |
gets options of a scaler | |
GAVL_PUBLIC int | gavl_video_scaler_init (gavl_video_scaler_t *scaler, const gavl_video_format_t *src_format, const gavl_video_format_t *dst_format) |
Initialize a video scaler. | |
GAVL_PUBLIC int | gavl_video_scaler_init_convolve (gavl_video_scaler_t *scaler, const gavl_video_format_t *format, int h_radius, const float *h_coeffs, int v_radius, const float *v_coeffs) |
Initialize a video scaler as a generic convolver. | |
GAVL_PUBLIC void | gavl_video_scaler_scale (gavl_video_scaler_t *scaler, const gavl_video_frame_t *input_frame, gavl_video_frame_t *output_frame) |
Scale video. |
Video scaler.
The scaler does the elementary operation to take a rectangular area of the source image and scale it into a specified rectangular area of the destination image. The source rectangle has floating point coordinates, the destination rectangle must have integer coordinates, which are aligned to chroma subsampling factors.
The internal scale tables are created for each plane and field separately. This means that:
You can use the scaler directly (through gavl_video_scaler_t). The generic video converter (gavl_video_converter_t) will create an internal scaler if necessary.
typedef struct gavl_video_scaler_s gavl_video_scaler_t |
Opaque scaler structure.
You don't want to know what's inside.
GAVL_PUBLIC gavl_video_scaler_t* gavl_video_scaler_create | ( | ) |
Create a video scaler.
GAVL_PUBLIC void gavl_video_scaler_destroy | ( | gavl_video_scaler_t * | scaler | ) |
Destroy a video scaler.
scaler | A video scaler |
GAVL_PUBLIC gavl_video_options_t* gavl_video_scaler_get_options | ( | gavl_video_scaler_t * | scaler | ) |
gets options of a scaler
scaler | A video scaler |
After you called this, you can use the gavl_video_options_set_*() functions to change the options. Options will become valid with the next call to gavl_video_scaler_init
GAVL_PUBLIC int gavl_video_scaler_init | ( | gavl_video_scaler_t * | scaler, | |
const gavl_video_format_t * | src_format, | |||
const gavl_video_format_t * | dst_format | |||
) |
Initialize a video scaler.
scaler | A video scaler | |
src_format | Input format | |
dst_format | Output format |
You should have equal pixelformats in the source and destination. This function can be called multiple times with one instance.
GAVL_PUBLIC int gavl_video_scaler_init_convolve | ( | gavl_video_scaler_t * | scaler, | |
const gavl_video_format_t * | format, | |||
int | h_radius, | |||
const float * | h_coeffs, | |||
int | v_radius, | |||
const float * | v_coeffs | |||
) |
Initialize a video scaler as a generic convolver.
scaler | A video scaler | |
format | Format (must be the same for input and output) | |
h_radius | Horizontal radius | |
h_coeffs | Horizontal coefficients | |
v_radius | Vertical radius | |
v_coeffs | Vertical coefficients |
This initialized a scaler for use as a generic convolver. The h_radius and v_radius arguments denote the numbers of pixels, which are taken in both left and right from the center pixel, i.e. a value of 1 will result in a 3-tap filter. The coefficients must be given for ALL taps (the convolver does not assume the coeffitients to be symmetric)
This function can be called multiple times with one instance.
GAVL_PUBLIC void gavl_video_scaler_scale | ( | gavl_video_scaler_t * | scaler, | |
const gavl_video_frame_t * | input_frame, | |||
gavl_video_frame_t * | output_frame | |||
) |
Scale video.
scaler | A video scaler | |
input_frame | Input frame | |
output_frame | Output frame |