白丝美女被狂躁免费视频网站,500av导航大全精品,yw.193.cnc爆乳尤物未满,97se亚洲综合色区,аⅴ天堂中文在线网官网

Handwriting detector, extractor, and language classifier

專利號(hào)
US11176361B2
公開日期
2021-11-16
申請(qǐng)人
Raytheon Company(US MA Waltham)
發(fā)明人
Darrell L. Young; Kevin C. Holley
IPC分類
G06F40/171; G06F40/263; G06K9/00; G06K9/34; G06K9/38; G06K9/62; G06K9/68; G06K9/72
技術(shù)領(lǐng)域
language,may,or,in,bounding,be,hardware,features,geometric,image
地域: MA MA Waltham

摘要

Disclosed are methods for handwriting recognition. In some aspects, an image representing a page of a sample document is analyzed to identify a region having indications of handwriting. The region is analyzed to determine frequencies of a plurality of geometric features within the region. The frequencies may be compared to profiles or histograms of known language types, to determine if there are similarities between the frequencies in the sample document relative to those of the known language types. In some aspects, machine learning may be used to characterize the document as a particular language type based on the frequencies of the geometric features.

說(shuō)明書

Operation 2020 determines a count of geometric features occurring along the length of the region. For example, operation 2020 may analyze the region starting at a first side (defined along the length) and completing at a second side of the region (defined along the length). Operation 2020 may count how many of each type of geometric features is detected within the region.

In operation 2030, the counts are normalized based on the height. For example, as discussed above with respect to process 1800, the counts may be normalized according to equation 2 to define frequencies. Operation 2030 operates to adjust for different scales of handwriting.

In operation 2040, the normalized counts are provided to a trained model. For example, as discussed above with respect to FIG. 19, information 1990 is provided to a feature determination module 1950b, which may determine the normalized counts discussed above with respect to operations 2020 and 2030.

In operation 2050, a handwriting language type included in the region is determined. The language type is determined by the trained model based on the normalized counts. As discussed above, the normalized counts define frequencies of certain geometric features within the region. The geometric features may include one or more of loops, rectangles, squares, boxes, curves, orthogonal intersections, cross-overs, corners, closed curves, or connected curves.

權(quán)利要求

1
微信群二維碼
意見反饋