Batch prediction

[aniketmaurya]

Hi! Is there any plan for batch inference or passing list of images/image path list in the readtext function?

I see that recognize takes batch_size arg but img_cv_grey seem to be single image only. How do I infer on a batch of image?

def recognize(self, img_cv_grey, horizontal_list=None, free_list=None,\
                  decoder = 'greedy', beamWidth= 5, batch_size = 1,\
                  workers = 0, allowlist = None, blocklist = None, detail = 1,\
                  rotation_info = None,\
                  paragraph = False,\
                  contrast_ths = 0.1,adjust_contrast = 0.5, filter_ths = 0.003,\
                  reformat=True):

[rkcosmos]

It is currently processing images one-by-one. Adjusting batch_size may improve your GPU utilization. This means EasyOCR will process several text boxes (in the same images) at the same times.

[SamSamhuns]

I have opened a PR #458 for this

[aniketmaurya]

awesome! Thank you @SamSamhuns