Function: dvpp_vpc_batch_crop_resize_paste_async
C Prototype |
aclError acldvppVpcBatchCropResizePasteAsync(acldvppChannelDesc *channelDesc, acldvppBatchPicDesc *srcBatchPicDescs, uint32_t *roiNums, uint32_t size, acldvppBatchPicDesc *dstBatchPicDescs, acldvppRoiConfig *cropAreas[], acldvppRoiConfig *pasteAreas[], acldvppResizeConfig *resizeConfig, aclrtStream stream) |
|---|---|
Python Function |
ret = acl.media.dvpp_vpc_batch_crop_resize_paste_async(dvpp_channel_desc, src_batch_pic_descs, roi_nums_list, dst_batch_pic_descs, crop_area_list, paste_area_list, resize_config, stream) |
Function Usage |
Crops one or more ROIs from one or more images and pastes each cropped image to the canvas for output. This API is asynchronous. A cropped image will be resized again if crop_area and paste_area have different sizes. The resizing algorithm specified in resize_config is used. |
Input Description |
dvpp_channel_desc: int, pointer address of the channel description. It must be the same as the dvpp_channel_desc argument passed to the acl.media.dvpp_create_channel call. src_batch_pic_descs: int, pointer address of the description of batch input images.
roi_nums_list: list, array for the number of crop ROIs, up to 256. The total value is less than or equal to 256, which is the same as the value of batch_size in dst_batch_pic_descs. (roi_nums_list[0] + ... + roi_nums_list[size - 1] <= 256) dst_batch_pic_descs: int, pointer address of the description of batch output images.
crop_area_list: list, list of the position of the cropping area.
paste_area_list: list, list of the position of the paste area.
resize_config: int, pointer address of the image resizing configuration data. Call acl.media.dvpp_create_resize_config to create an image resizing configuration in advance. stream: int, stream. |
Return Value |
ret: int, error code.
|
Restrictions
Version |
Restriction |
|---|---|
Atlas 200/300/500 Inference Product Atlas Training Series Product |
|