devjeonghwan
devjeonghwan
It depends on where the memory of the image to be encoded is located. if the image memory is in the CUDA(Device), AV_PIX_FMT_CUDA is faster (It can remove the bottleneck...
Measure and compare image processing and encoding performance This method can help. 1. Separate the image processing thread and encoder thread. 2. Make the encoder thread poll the queue for...
Create and initialize `AVFilterContext` using `nppscale_init` and then define `in`, `out` frame informations(pixel format, size, etc..) to `NPPScaleStageContext`. Now, you can call npp resize function like this. ``` static int...
Can you upload "hs_err_pid" file?
The javacpp preset for Tesseract is already "5.0.1" in "1.5.7" tag In your project POM, change javacpp version to 1.5.7 and tersseract version to 5.0.1. ```xml 4.0.0 org.bytedeco.tesseract BasicExample 1.5.7...
Same problem for a year...
This is likely related to this FFmpeg [issue](https://trac.ffmpeg.org/ticket/11012). What versions of JavaCPP and FFmpeg preset are you using?
Can you share your APNG file for testing?
This is an FFmpeg bug. There is no proper `tEXt` parsing logic for APNG decoding, and the format assumptions are incorrect. You should create an issue in the FFmpeg repository...
Unfortunately, I'm not an FFmpeg contributor, so I can't provide definitive guidance. However, manipulating the APNG binary isn't particularly difficult. Please check the code below: ```java public static byte[] removeApngTextChunks(byte[]...