RGB888_U8 to RGB (or BGR)

Assume that AIPP inputs an RGB888_U8 image and outputs an RGB image. The input image size is 250 x 250, and the valid data region starts from the (0, 0) pixel coordinates in the upper-left corner. For the original network model, C is 3, and H and W are 240.

• Enable the cropping (crop) function.
• Set the horizontal and vertical coordinates (load_start_pos_h and load_start_pos_w) of the cropping start position to 0.
• Do not need to configure the crop_size_w and crop_size_h parameters. In this case, the width and height of the cropped image size (crop_size[W|H]) are obtained from the width and height of the --input_shape option used during model conversion, and the 240 x 240 region is selected starting from the (0, 0) coordinates.
• Do not need to configure the rbuv_swap_switch, csc_switch, and corresponding CSC matrix parameters.

aipp_op {
    aipp_mode: static
    input_format: RGB888_U8
    csc_switch: false
    related_input_rank: 0
    src_image_size_w: 250
    src_image_size_h: 250
    crop: true
    load_start_pos_w: 0
    load_start_pos_h: 0
    # If the original model requires the BGR format, set rbuv_swap_switch to true.
    rbuv_swap_switch: false
    # The normalization parameters are set as required. In this sample, these parameters are set to default values, so the pixel values are not changed.
    # The configured normalization parameters are applied to channels after channel swap.
}