experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix expensive decompression in JPEG GetImageInfo

Browse Source 
The current implementation of GetImageInfo starts a libjpeg
decompression pass over input images. Certain images
(e.g., progressive JPEG) trigger a full image decompression, resulting
in a performance degradation. This commit switches to an equivalent but
cheaper libjpeg call for evaluating image dimensions.
This commit is contained in:
Michael Kuchnik 2020-10-15 17:32:28 -04:00
parent d6e0181f1d
commit c5a6556942
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tensorflow/core/lib/jpeg/jpeg_mem.cc
View File

@ -577,7 +577,7 @@ bool GetImageInfo(const void* srcdata, int datasize, int* width, int* height,
SetSrc(&cinfo, srcdata, datasize, false);
jpeg_read_header(&cinfo, TRUE);
jpeg_start_decompress(&cinfo); // required to transfer image size to cinfo
jpeg_calc_output_dimensions(&cinfo);
if (width) *width = cinfo.output_width;
if (height) *height = cinfo.output_height;
if (components) *components = cinfo.output_components;