picture text recognition software for chinese

#Picture text recognition software for chinese full#

As Kersten (1987) pointed out, humans could make reliable perceptual inferences in this case because objects in the world are shaped in such a way that their images carry redundant or correlated luminance information between neighboring regions. For example, using a partially occluded face image, Kersten (1987) demonstrated that humans could use the luminance of the face region that surrounds a small occluder to infer the luminance of the face behind the occluder. In this case, what a perceiver can do is infer the missing information from the regularities of objects in the world. However, this completion cannot be completely certain. One might expect that the visual system, working from the stimulus input, first “perceptually completes” the shape rather than encoding the raw stimulus image as is. This question is particularly important when the stimulus is impoverished, for example, when partially occluded or imbedded in noise, because imperfect stimuli are common-place. One important question in human visual perception is how the visual system encodes an input stimulus into memory. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.ĭata Availability: Data are deposited into Figshare (DOI: ).įunding: This work was supported by the following sources of funding: ZL: US National Science Foundation (NSF)(BCS 9817979) ZL, YZ: China Natural National Science Foundation (NSF)(NSFC 31228009)(NSFC 31230032).Ĭompeting interests: The authors have declared that no competing interests exist. Received: DecemAccepted: Published: June 3, 2016Ĭopyright: © 2016 Chang et al.

PLoS ONE 11(6):Įditor: Philip Allen, University of Akron, UNITED STATES We conclude that top-down influence from character recognition was present but very limited, at least with the task and stimuli used.Ĭitation: Chang J, Zhou Y, Liu Z (2016) Limited Top-Down Influence from Recognition to Same-Different Matching of Chinese Characters. This small effect was further confirmed with native speakers in China, who should give rise to the largest possible effect.

However, the magnitude of this effect was small, to the extent that the average same-different accuracies were comparable for readers and non-readers alike. The only indication of top-down influence was the readers’ higher accuracy in matching upright than inverted and reflected characters. Interestingly, although the readers’ recognition accuracies well correlated with their years of Chinese language schooling, they were uncorrelated with the matching accuracies in the same-different task with upright characters. The participants who could recognize some characters (defined as readers) were subsequently tested with character recognition in a naming task.

#Picture text recognition software for chinese full#

The participants’ Chinese reading fluency spanned the full range, from not knowing any characters to native speakers. The two characters were either both upright or both inverted and mirror-reflected. In each experimental trial, two partially occluded characters were shown sequentially, and participants decided whether or not they were the same. We investigated the extent to which recognition of Chinese characters influenced same-different matching performance that did not require recognition.