Method for generating word code, method and device for recognizing codes

A method for generating word code, comprising:

• acquiring target words and a corresponding target file which are inputted by a user;
• splitting each word in the target words on the basis of stroke connecting points to obtain split sequences;
• determining whether the number of the split sequences is sufficient to express characters in the target file;
• if the number of the split sequences is sufficient to express the characters, generating word codes according to a preset encoding mechanism:
• selecting at least part of the split sequences;
• performing visual anomaly encoding processing on various attribute values of the at least part of the split sequences to obtain machine visual anomaly features, wherein the machine visual anomaly features are used for directly expressing the characters of the target file, and the machine visual anomaly features corresponds to the characters of the target file;
• if the number of the split sequences is not sufficient to express the characters, randomly selecting, from the split sequences, several split sequence elements to perform visual anomaly processing on the attribute values;
• sequentially recording sequence numbers of split sequence elements which are visual anomaly processed, the recorded sequence numbers are records as first sequence numbers;
• splicing pronunciations or acronyms of the words corresponding to the target words with the random numbers, so as to generate second sequence numbers;
• generating word codes according to the first sequence numbers and second sequence numbers, and creating the relationship between the word codes and the target file.

A method for recognizing word code, comprising:

• acquiring an image containing word codes, and recognizing target words corresponding to the word codes;
• splitting each word in the target words on the basis of stroke connecting points, so as to obtain word code split sequences;
• recognizing visual anomaly sequences from the word code split sequences;
• and
• determining whether the word codes are generated on the basis of an encoding mechanism designed by a system, and if the word codes are generated on the basis of an encoding mechanism designed by a system, directly reading the word codes and performing target file invoking;
• if the word codes are not generated on the basis of an encoding mechanism designed by a system, invoking a preset target file corresponding to the word codes according to the first sequence numbers and second sequence numbers, which are corresponding to the target words;
• wherein the first sequence numbers and second sequence numbers are determined through the following process:
• sequentially recording sequence numbers of split sequence elements which are visual anomaly processed, the recorded sequence numbers are records as the first sequence numbers;
• splicing pronunciations or acronyms of the words corresponding to the target words with the random numbers, so as to generate second sequence numbers.

A device for recognizing word code, comprising:

• a target words recognition module, configured to acquire an image containing word codes, recognize target words corresponding to the word codes, and trigger a word code splitting module;
• the word code splitting module, configured to split each word in the target words on the basis of stroke connecting points, so as to obtain word code split sequences, and trigger an visual anomaly recognition module;
• the visual anomaly recognition module, configured to recognize first sequence number from the word code split sequences and trigger a target file invoking module;
• and
• the target file invoking module, configured to determine whether the word codes are generated on the basis of an encoding mechanism designed by a system, and if the word codes are generated on the basis of an encoding mechanism designed by a system, directly reading same, and performing target file invoking;
• the word codes are not generated on the basis of an encoding mechanism designed by a system, invoking a preset target file corresponding to the word codes according to the first sequence numbers and second sequence numbers, which are corresponding to the target words;
• wherein the first sequence numbers and second sequence numbers are determined through the following process:
• sequentially recording sequence numbers of split sequence elements which are visual anomaly processed, the recorded sequence numbers are records as the first sequence numbers;
• splicing pronunciations or acronyms of the words corresponding to the target words with the random numbers, so as to generate second sequence numbers.