Information offering method and information offering device

[0040] Embodiment 1: Recognize the text in the image, and determine a participle composed of the recognized text, and when the determined participle contains a preset keyword, determine the participle as the recognized content information.

[0041] Continuing with the above example, assuming that the preset keywords include "park", the server recognizes the text in the image and determines that the participle "Boda Park" composed of the recognized text contains the preset keyword "park", so the " Boda Park" is determined as the identified content information.

[0042] Embodiment 2: Recognize the text in the image, and determine the participle composed of the recognized text. When the determined participle contains preset keywords, and the text after the determined participle is a specified type of text, the specified The text of the type is determined as the recognized content information.

[0043] For example, assume that the user takes a photo, which contains the content information of the name of the restaurant, and the name of the restaurant is "302, a roast duck restaurant in Qianmen". Assuming that the content information to be identified is the content information of the branch serial number "302", the method of identifying the serial number is as follows:

[0044] If the default keyword is set to "Roast Duck Restaurant", and the specified type of text is set to a numeric type of text, the server will recognize the text in the image and determine that the participle "Qianmen Roast Duck Restaurant" composed of the recognized text

[0045]Embodiment 3: Recognize the text in the image, and determine the participle composed of the recognized text, when the recognized text is a specified type of text, and it is determined that the participles composed of the specified type of text fall into the preset set , determine the participle composed of the specified type of text as the recognized content information.

[0046] In actual application scenarios, the content information in some images may contain date-type content information. Assuming that the image contains the content information "May 28, 2014", identify the content information "May 28, 2014", as follows:

[0047] For example, set the text of the specified type to the text of the numeric type, and set the default set to include: the first set of 15 word segmentations from 2000 to 2014, the second set of 12 word segmentations from 01 to 12, and the second set of 12 word segmentations from 01 to 31 The third set of 31 word segmentations.

[0048] Then the