Method for rendering subtitling based on curved profile closed loop domain and pixel mask matrix

This problem has brought many disadvantages to the subtitle system in both implementation and application

From an implementation point of view, this will lead to multiple completely different rendering engines and rendering processes in a subtitle system, and if you want to support a new subtitle object, you may have to define and implement a new rendering engine, resulting in subtitles There are great limitations in the functional scalability of the system; from the application point of view, it brings two fatal problems: on the one hand, different types of subtitle objects will eventually display inconsistent effects, for example, two different types of subtitle objects have the same It is a full edge with two pixels added, but due to different rendering engines, the final processing results may be different, which brings a lot of inconvenience to users when creating subtitles; on the other hand, in the subtitle system for The operation of subtitle objects has to be subject to specific restrictions. For example, some subtitle objects support geometric rotation and geometric scaling, while others do not. Some subtitle objects support adding edges and shadows at any angle, while others only support fixed Adding edges and shadows to angles, some don’t even support it

This will lead to two problems. On the one hand, since the final processing result requires an integer pixel value, rounding operations will be performed at different stages of the rendering engine algorithm, which will cause cumulative errors in calculation, resulting in the final The subtitle effect has defects in local details, such as blurred text with many strokes, sticky strokes, etc.; on the other hand, these floating-point operations are mathematical operations that take up a lot of machine instructions, which greatly reduces the performance of the entire subtitle system. Work efficiency affects the timeliness of subtitle production and broadcasting

[0010] Third, the effect of edge anti-aliasing, which is one of the important indicators of subtitle rendering, is not ideal, and cannot even reach a practical level.

This is because the existing subtitle system completely relies on the image processing method of the current pixel space when performing anti-aliasing processing. Without the participation of subtitle geometric vector information, sub-pixel-based processing cannot be realized.

This makes the final rendering effect of the existing subtitle system have a relatively obvious edge aliasing effect, which affects the effect of the subtitle display

[0011] Fourth, for the same reason as above, when the image processing of the subtitle object based on the FilterPipe (FilterPipe) is performed, the final effect of the subtitle will also be affected

However, since there is no participation of subtitle geometric vector information, the processing of the filter pipeline will destroy the anti-aliasing effect of the former, that is to say, new edge aliasing effects will be accumulated on the edge processing that is not ideal.

[0012] Fifth, when a subtitle object performs gradient color rendering, it should perform radial gradient color processing according to the vector outline of the object, but in the existing subtitle system, the gradient color rendering process is to put the gradient color into In a memory block, when rendering, the corresponding pixel map method is still adopted

This makes the existing subtitle system unable to achieve the rendering effect of gradient color in the true sense

[0013] Sixth, the existing subtitle system cannot add multiple outer edges, multiple inner edges, multiple side three-dimensional edges, and multiple shadows

From the above analysis of the application characteristics of the subtitle system, it can be concluded that the current TV program production has higher and higher requirements for the subtitle system, and the simple surface, edge, and shadow superimposed subtitle display can no longer meet this application demand.

[0014] Seventh, the rendering of the existing subtitle system is not based on the Bezier curve outline, but since the fonts in the TrueType font library are described by the Bezier curve, it is necessary to convert the Bezier curve extracted from the font library into another In the form of vector, due to the asymmetry of information between different vector expressions, there will be a loss of information during this conversion process, which will deform the shape of the strokes, resulting in uneven thickness of the strokes, which will affect the appearance of the font.

[0016] Ninth, the existing subtitle system cannot convert text subtitle objects into graphic subtitle objects, so that functions such as character deformation, character reflection, character reversal, curve placement, and word art cannot be realized, and cannot meet the needs of advanced subtitle applications

[0017] Tenth, the existing subtitle system cannot implement many different types of bordering methods, such as sharp corners, rounded corners, square corners, etc.

This affects the wide application of subtitles to a certain extent

[0018] Eleventh, the existing subtitle system cannot support irregular graphics of any shape, so it cannot cope with the increasingly important types of program production such as geographic information description, weather forecast, and comprehensive information