diff options
Diffstat (limited to 'Software/Beignet')
| -rw-r--r-- | Software/Beignet/howto/video-motion-estimation-howto.mdwn | 20 |
1 files changed, 10 insertions, 10 deletions
diff --git a/Software/Beignet/howto/video-motion-estimation-howto.mdwn b/Software/Beignet/howto/video-motion-estimation-howto.mdwn index 829ba29d..6b99d54c 100644 --- a/Software/Beignet/howto/video-motion-estimation-howto.mdwn +++ b/Software/Beignet/howto/video-motion-estimation-howto.mdwn @@ -33,16 +33,16 @@ the steps as below: - The prototype of built-in kernel block_motion_estimate_intel is as following: - _kernel void - block_motion_estimate_intel - ( - accelerator_intel_t accelerator, - __read_only image2d_t src_image, - __read_only image2d_t ref_image, - __global short2 * prediction_motion_vector_buffer, - __global short2 * motion_vector_buffer, - __global ushort * residuals - ); + _kernel void + block_motion_estimate_intel + ( + accelerator_intel_t accelerator, + __read_only image2d_t src_image, + __read_only image2d_t ref_image, + __global short2 * prediction_motion_vector_buffer, + __global short2 * motion_vector_buffer, + __global ushort * residuals + ); So you should create related objects and setup these kernel arguments by clSetKernelArg. Create source and reference image object, on which you want to do video motion estimation. The image_channel_order should be CL_R and image_channel_data_type should be CL_UNORM_INT8. |