Jetson_multimedia_api jpeg encode encoding Issue on Jetson AGX orin JP 5.1.3

Autonomous Machines Jetson & Embedded Systems Jetson AGX Orin
,
1

Hi,
I added the continuous yuv data jpeg encoding function in the 05_jpeg_encode sample of Jetson_Multimedia_API_R35.5.0_aarch64.tbz2. The code is in the attachment 05_jpeg_encode.zip
05_jpeg_encode.zip (6.5 MB). I found that only the first frame can be encoded correctly, and all subsequent encodings are wrong. Execute the command ./jpeg_encode yuv420_1920_1536.yuv 1920 1536 test.jpg --encode-buffer --perf
It prints as follows:

pegasus@pegasus-ubuntu-3:~/workspace/jetson_multimedia/jetson_multimedia_api_v513/samples/05_jpeg_encode$ ./jpeg_encode yuv420_1920_1536.yuv 1920 1536 test.jpg --encode-buffer --perf

NvMMLiteBlockCreate : Block : BlockType = 1
out_file_name: output_1920_1536_0.jpg out_size: 187012
out_file_name: output_1920_1536_1.jpg out_size: 18446744069414771811
out_file_name: output_1920_1536_2.jpg out_size: 18446744069414771729
out_file_name: output_1920_1536_3.jpg out_size: 18446744065119804708
Could not read a complete frame from file
----------- Element = jpenenc -----------
Total units processed = 4
Average latency(usec) = 7221
Minimum latency(usec) = 3512
Maximum latency(usec) = 18244

App run failed

The data length from the second frame is wrong，
Can out suggest a workaround for this issue ?

Thanks
Liuyue

3

Hi,
For information, do yo observe the issue when calling encodeFromFd()? If it is specific to calling encodeFromBuffer(), we would suggest call encodeFromFd() as a quick solution.

4

Hi,
DaneLLL,Thank you very much for your reply,I have used the encodeFromFd function to replace the encodeFromBuffer function，Modify the code as
05_jpeg_encode.zip (5.9 MB)
，Execute the command ./jpeg_encode yuv420_1920_1536.yuv 1920 1536 test.jpg --perf. The problem still exists. The performance is the same. The first frame is good, but all subsequent frames have errors.

6

Hi,
Please reset the buffer size before calling encodeFromFd() or encodeFromBuffer():

out_buf_size = ctx.in_width * ctx.in_height * 3 / 2;
7

Hi,
DaneLLL,Thank you very much for your reply,This modification has solved the problem.

Thanks
Liuyue

9

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.