CN116051381A - A method of combining rendering blocks based on OpenImageIO - Google Patents

Abstract

The invention discloses a rendering block merging method based on OpenImageIO, comprising the following steps: S1: using OpenImageIO to read the block image to be merged; S2: reading the full_width, full_height and pixel value of the block image; S3: judging Whether the full_width and full_height after the merging of the current block picture are smaller than the full_width and full_height of the picture before the block; The pixels of a block image are spliced according to the first dimension to obtain a pixel array; S6: Create the same image.space blank image as the block image; S7: Write the spliced pixel array in the image.space blank image; S8: Output final image. The invention is applicable to images in various formats; it can reduce the excessive memory resources occupied in the synthesis process; reduce the complexity of the synthesis operation; and solve the problem of channel loss or channel synthesis damage.

Description

A method of combining rendering blocks based on OpenImageIO

technical field

The present invention relates to the technical field of cloud rendering, in particular to a method for merging rendering blocks based on OpenImageIO.

Background technique

Block rendering is a rendering method often used from 3D modeling to 2D images. Its purpose is to divide a complete image to be rendered into multiple blocks, and then each machine renders one of the blocks to achieve multi-machine A way to speed up rendering by co-rendering. However, the block-rendered image finally needs to be block-merged, and the images are stitched together according to the corresponding positions to obtain the final image result. Due to the various image formats and channels rendered in the fields of film and television rendering and 3D modeling rendering, and the parameters are complex, this brings difficulties to the merger. The current common block merging methods include: using the merging function provided by the rendering software itself; using professional image merging software such as Nuke; other image processing software based on open protocols.

1. At present, the main problem encountered in the block merging of rendered images is the merging problem caused by various image formats and parameters.

2. There are some major defects in the current common block merging method. For example, the merging function of the CG software itself often has unsupported formats, image damage after merging, long synthesis time, and large computer memory usage; for example If professional image processing software Nuke is used, the operation may be complicated, the synthesis result channel may be damaged, Chinese language is not supported, and some formats are not supported, and Nuke’s memory usage is also very high; other tools based on the development protocol are Generally, only certain common formats can be processed.

Therefore, there are defects in the prior art and need to be improved.

Contents of the invention

The technical problem to be solved by the present invention is to provide a method for merging rendering blocks based on OpenImageIO, which facilitates merging of images rendered by blocks.

The technical solution of the present invention is as follows: a method for merging rendering blocks implemented based on OpenImageIO is provided, including the following steps.

S1: Use OpenImageIO to read the block images to be merged.

S2: Read the full_width, full_height and pixel values of the block image.

S3: Determine whether the full_width and full_height of the merged picture of the current block are smaller than the full_width and full_height of the picture before the block; if yes; go to step S4; if not, go to step S5.

S4: Perform edge padding on each block image.

S5: splicing the pixels of each block image according to the first dimension to obtain a pixel array.

S6: Create the same image.space blank image as the tiled image.

S7: Write the spliced pixel array in the image.space blank image.

S8: Outputting the final image.

Obtain the pixels of each block image, fill the edges of the required block images, and then stitch all the pixels of the block images to obtain a pixel array, and write the pixel array in the image.space blank image, so as to obtain the final The rendered image of . Since the pixels of the block image are obtained for splicing, it is compatible with pictures of various formats, which greatly facilitates the application.

The step S1 is: use OpenImageIO to read the segmented pictures to be merged to obtain image.space and rgb pixel values. In step S1, the file object pointer of the block image is read by using the method of reading the file cache image; the subsequent reading reads the file by continuously moving the pointer. In the subsequent reading, the purpose of reading files with low resources is achieved by continuously moving the pointer.

The step S2 is: read the image format, resolution, full image width and height, channel number, pixel value, channel name and whether it contains a depth image through image.space.

The filling rule in the step S4 is as follows: for the height, the filling value is: (tile_spec.full_height/tile_count)-tile_spec.height; for the width, the filling value is: (tile_spec.full_width/tile_count)-tile_spec.width-tile_spec. x; where tile_spec.full_height represents the height of the entire image, tile_spec.full_width represents the height of the entire image, tile_spec.height represents the height of the block image, tile_spec.width represents the height of the block image, and tile_spec.x represents the real pixel range of the block image , tile_count indicates the number of blocks.

In step S5, the pixels of each block image are concatenate according to the first dimension.

Said step S7 is: all block image parameters use the data read in step S2, modify the full_width and full_height of the block image to be the width and height of the picture after synthesis, and write the multi-dimensional image after splicing into the image.space blank image pixel array.

By adopting the above scheme, the present invention provides a rendering block merging method based on OpenImageIO, which has the following technical effects:

1. Suitable for images in various formats, such as BMP, Cineon, DDS, DICOM, DPX, Field3D, FITS, GIF, HDR/RGBE, HEIF/FEIC/AVIF, ICO, IFF, JPEG, JEPG-2000, Movie formats, Null format, OpenEXR, OpenVDB, PNG, PNM/Netpbm, PSD, Ptex, Raw digital camera files, RLA, SGI, Softimage PIC, Target, Term(Termianl), TIFF, Webbp, Zfile;

2. Reduce the excessive memory resources occupied during the synthesis process;

3. Reduce the complexity of the synthesis operation;

4. Solve the problem of channel loss or channel synthesis damage.

Description of drawings

Fig. 1 is a flow chart of the method of the present invention.

Detailed ways

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

Referring to FIG. 1 , this embodiment provides a method for merging rendering blocks based on OpenImageIO, including the following steps.

S1: Use OpenImageIO to read the segmented images to be merged, and obtain image.space and rgb pixel values. In step S1, the file object pointer of the block image is read by using the method of reading the file cache image; the subsequent reading reads the file by continuously moving the pointer. In the subsequent reading, the purpose of reading files with low resources is achieved by continuously moving the pointer.

S2: Read the image format, resolution, full image width and height, channel number, pixel value, channel name and whether to include depth image of the block image through image.space.

S3: Determine whether the full_width and full_height of the merged picture of the current block are smaller than the full_width and full_height of the picture before the block; if yes; go to step S4; if not, go to step S5.

S4: Perform edge padding on each block image.

S5: Concatenate the pixels of each block image according to the first dimension to obtain a pixel array.

S6: Create the same image.space blank image as the tiled image.

S7: Use the data read in step S2 for all block image parameters, modify the full_width and full_height of the block image to be the width and height of the synthesized image, and write the spliced multi-dimensional pixel array into the image.space blank image.

S8: Outputting the final image.

The filling rule in the step S4 is as follows: for the height, the filling value is: (tile_spec.full_height/tile_count)-tile_spec.height; for the width, the filling value is: (tile_spec.full_width/tile_count)-tile_spec.width-tile_spec. x; where tile_spec.full_height represents the height of the entire image, tile_spec.full_width represents the height of the entire image, tile_spec.height represents the height of the block image, tile_spec.width represents the height of the block image, and tile_spec.x represents the real pixel range of the block image , tile_count indicates the number of blocks.

In summary, the present invention provides a method for merging rendering blocks based on OpenImageIO, which has the following technical effects:

1. Suitable for images in various formats, such as BMP, Cineon, DDS, DICOM, DPX, Field3D, FITS, GIF, HDR/RGBE, HEIF/FEIC/AVIF, ICO, IFF, JPEG, JEPG-2000, Movie formats, Null format, OpenEXR, OpenVDB, PNG, PNM/Netpbm, PSD, Ptex, Raw digital camera files, RLA, SGI, Softimage PIC, Target, Term(Termianl), TIFF, Webbp, Zfile;

2. Reduce the excessive memory resources occupied during the synthesis process;

3. Reduce the complexity of the synthesis operation;

4. Solve the problem of channel loss or channel synthesis damage.

The above are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention. Inside.

Claims (7)

1. A rendering block merging method realized based on OpenImageIO, is characterized in that, comprises the following steps: S1: Use OpenImageIO to read the block images to be merged; S2: Read the full_width, full_height and pixel values of the block image; S3: Judging whether the full_width and full_height of the merged picture of the current block are smaller than the full_width and full_height of the picture before the block; if so; go to step S4: if not, go to step S5: S4: fill the edge of each block picture; S5: splicing the pixels of each block image according to the first dimension to obtain a pixel array; S6: Create the same image.space blank image as the block image; S7: Write the spliced pixel array in the image.space blank image; S8: Outputting the final image. 2. A method for merging rendering blocks based on OpenImageIO according to claim 1, wherein said step S1 is: using OpenImageIO to read the block images to be combined to obtain image.space and rgb pixel values. 3. A kind of method for merging rendering blocks realized based on OpenImageIO according to claim 2, characterized in that, said step S2 is: read the image format, resolution, and full image of the block image through image.space Width and height, number of channels, pixel value, channel name and whether to include a depth image. 4. a kind of rendering block merging method based on OpenImageIO according to claim 1 is characterized in that, in step S1, uses the mode of reading file cache image, reads the file object pointer of block image; Subsequent reads read the file by continuously moving the pointer. 5. A kind of rendering block merging method realized based on OpenImageIO according to claim 1, characterized in that, the filling rules in the step S4 are as follows: For height, the filling value is: (tile_spec.full_height/tile_count)-tile_spec.height; For width, the padding value is: (tile_spec.full_width/tile_count)-tile_spec.width-tile_spec.x; Among them, tile_spec.full_height represents the height of the entire image, tile_spec.full_width represents the height of the entire image, tile_spec.height represents the height of the block image, tile_spec.width represents the height of the block image, tile_spec.x represents the real pixel range of the block image, and tile_count represents the number of blocks. 6. A method for merging rendering blocks based on OpenImageIO according to claim 1, characterized in that in step S5, the pixels of each block image are concatenate according to the first dimension. 7. A kind of method of rendering block merging based on OpenImageIO according to claim 1, characterized in that, said step S7 is: all block image parameters use the data read in step S2 to modify the block The full_width and full_height of the image are the width and height of the composited image, and the blank image in image.space is written into the spliced multi-dimensional pixel array.

Images