Google.cn license

It can't be all open. Others like Beijing Forestry University student Zhang Meng, 22, said one major reason he uses Google is because it offers uncensored information. But other Internet users say Google's presence in China is still limited. A former Google user, Wang himself switched over to Baidu because his friends and customers all use it. Coinciding with the operating license issue is Google not getting a place on a Chinese government list of companies that will be able to offer Internet mapping services pending approval.

The State Bureau of Surveying and Mapping released a list naming 23 domestic companies to be granted a license to provide the online mapping services. Major companies like Baidu, search engine provider Sohu and e-commerce site Alibaba, made the list. In a statement, Google said, "China recently implemented a wide-ranging set of rules relating to online mapping. We are examining the regulations to understand their impact on our maps products in China.

Here are the latest Insider stories. More Insider Sign Out. Sign In Register. But other Internet users say Google's presence in China is still limited. A former Google user, Wang himself switched over to Baidu because his friends and customers all use it. Coinciding with the operating license issue is Google not getting a place on a Chinese government list of companies that will be able to offer Internet mapping services pending approval.

The State Bureau of Surveying and Mapping released a list naming 23 domestic companies to be granted a license to provide the online mapping services. Major companies like Baidu, search engine provider Sohu and e-commerce site Alibaba, made the list. In a statement, Google said, "China recently implemented a wide-ranging set of rules relating to online mapping. We are examining the regulations to understand their impact on our maps products in China. Here are the latest Insider stories.

More Insider Sign Out. Sign In Register. Sign Out Sign In Register. In order to improve the accuracy of the positions of the N character images, the embodiment shown in fig. Specifically, step S in fig. Step S and aiming at the second image, obtaining each connected domain contained in the second image and the first position of each connected domain according to a connected domain method, calculating the average width of the connected domains according to the connected domains and the first positions, and determining the average width as the average character width of the second image.

The connected domain method utilizes the characteristic that each character or part of the character forms a connected domain, cuts the connected domains out, namely cuts out each character image, and meanwhile, can obtain the first position of each connected domain. Step S and judging whether a first character image with the character image width larger than a preset first width threshold exists in the N character images to be corrected according to the average character width, and if so, executing the step S If not, the step S of the embodiment shown in fig.

Step S and correcting the first character image according to the first position, and determining N character images contained in the second image according to the corrected first character image. Specifically, the process of correcting the first character image is different for different situations, including the situation that the first and last characters of the license plate are adhered to the rivet or the license plate characters are embedded into each other, and in the two situations, the N character images have the situation that the characters are too wide.

In order to solve the problem of too wide character and improve the accuracy of dividing the character, in one embodiment of step S, that is, modifying the first character image according to the first position may include:. For example, the position of the first character image is [20, 30], then a target connected component, which may include at least two connected components, is determined from the connected components, with the character position also being [20, 30 ].

That is, the first character image is actually a composite of a rivet and an english alphabet, and the corresponding target connected field includes a rivet connected field and an english alphabet connected field. Step 2: and segmenting the first character image or re-determining the position of the first character image according to the target connected domain and the corresponding first position.

For example, the position of the first character image is [20, 30], the first positions of the target connected domain are [20, 25] and [25, 30], and this case belongs to the adhesion of two characters, or the existence of a stain, a stain and a character adhesion on the license plate. The position of the first character image can be corrected to [20, 25] and [25, 30], i.

If the first positions of the target connected domains are [20, 22] and [21, 30], which are the position of the first target connected domain and the position of the second target connected domain, respectively, this case belongs to rivet and character adhesion.

Since it can be determined from the average character width assumed to be 10 that the width of the first target connected component is too small width is 2 , the first target connected component can be discarded and the position of the second target connected component can be determined as the position of the first character image.

As another example, the first character image has a position [20, 30], the first positions of the target connected components have positions [20, 27] and [23, 30], which pertains to the case where characters are embedded in each other. Then, the positions of the first character images may be corrected to [20, 25] and [25, 30 ]. In the case where such a license plate is character-divided according to the vertical projection method, 7 and 4 are divided into one character, and thus 7 and 4 are considered to be embedded in each other.

This embodiment can segment such mutually embedded characters. As can be seen from the above, in this embodiment, the connected domain method is adopted to further correct the N segmented character images to be corrected, and the accuracy of the character segmentation result can be improved for the situations of character adhesion, inaccurate position of the first and last characters, and character embedding in practical application.

For example, the horizontal lines in the character H are highly reflective, resulting in the character segmentation process recognizing H as two I, i. Therefore, in order to further improve the accuracy of the character segmentation result, the embodiment shown in fig.

Step S and judging whether a second character image with the character image width smaller than a preset second width threshold exists in the N character images to be corrected, and if so, executing the step S Specifically, if the image does not exist, it is indicated that the character splitting phenomenon does not exist in the N character images to be corrected, step S is executed, that is, the step of recognizing the characters in the N character images according to the pre-stored character features of each country and region is executed, and obtaining the license plate number corresponding to the first image is executed.

Step S and determining at least two adjacent second character images as character images to be corrected. Step S and obtaining an edge image of the second image, performing character segmentation on the edge image according to a vertical projection method, and determining a target character image corresponding to the position of the character image to be corrected from the segmented edge image.

Step S and inputting the character image to be corrected and the corresponding target character image into a character classifier to obtain the confidence coefficient of each character image.

The character classifier is set according to character features of each country and region. The character features of each country and region can be learned from a large number of license plates comprising each country and region by using a machine learning algorithm.

Step S and judging whether the confidence degrees of the character images to be corrected are smaller than the confidence degree of the corresponding target character image, if so, executing the step S Specifically, if not, it is determined that the target character image is not completely authentic, and at this time, the step S may be continuously performed.

Step S and combining at least two second character images contained in the character image to be corrected according to the position of the corresponding target character image to obtain N character images contained in the second image.

For example, the position of the target character image K is [20, 30], the positions of two second character images a and B included in the character image to be corrected are [20, 23] and [26, 30], respectively, where the confidences of A, B and K are 0.

As can be seen from the above, in the embodiment, when the width of the character image in the N character images to be corrected is smaller than the preset second width threshold, the case that the character splitting exists in the N character images to be corrected is described. And performing character segmentation on the edge image of the second image, and correcting the N character images to be corrected according to the character images obtained from the edge image. The character segmentation is carried out on the edge image of the second image, so that the obtained segmentation result is possibly different from the segmentation result obtained by directly carrying out character segmentation on the second image, the edge image is less influenced by external factors such as uneven illumination, and the characters at the uneven illumination position on the license plate can be more accurately identified.

Step S and judging whether the second image belongs to a double-layer license plate or not according to the width-height ratio of the second image and a preset ratio threshold, and if so, executing the step S And if not, indicating that the second image does not belong to the double-layer license plate, executing a step of obtaining a vertical projection graph of pixel values of the second image according to a vertical projection method aiming at each second image, and segmenting the second image according to the vertical projection graph to obtain N character images contained in the second image.

In practical applications, a double-layer license plate may exist in the second image, for example, the third license plate in the chinese hong kong license plate in fig.

Specifically, the width of the license plate in the world is generally pixels, the height is pixels, and a preset proportion threshold value can be determined according to the information. Step S obtaining an upper locating block and a lower locating block contained in each second image from the second images, specifically comprising:. It can be understood that, for a double-layer license plate, a certain gap exists between the upper character portion and the lower character portion.

The pixel jump characteristic value of the character part is obviously larger than that of the gap part, so that the pixel rows with similar pixel jump characteristic values can be identified and the brother attribute can be set according to the preset second jump characteristic threshold.

Since the upper layer character part has a different sibling attribute from the lower layer character part, the upper locating block and the lower locating block included in the second image can be obtained according to the value of the sibling attribute. Step S and converting the second image into a single-layer license plate image according to the upper positioning block and the lower positioning block. Specifically, in step S, converting the second image into a single-layer license plate image according to the upper locating block and the lower locating block may include: placing the upper positioning block at the left side of the lower positioning block, and further converting the second image into a single-layer license plate image; may also include: and placing the lower positioning block on the left side of the upper positioning block, and further converting the second image into a single-layer license plate image.

Of course, there are many embodiments for converting the second image into a single-layer license plate image according to the upper locating block and the lower locating block. Step S and aiming at each second image after being converted into the single-layer license plate image, obtaining a vertical projection image of the pixel value of the second image according to a vertical projection method, and segmenting the second image according to the vertical projection image to obtain N character images contained in the second image.

Step S is the same as step S in the embodiment shown in fig. As can be seen from the above, in this embodiment, whether the second image belongs to a double-layer license plate can be determined according to the aspect ratio of the second image. When the second image belongs to a double-layer license plate, the double-layer license plate can be converted into a single-layer license plate according to a certain interval between an upper layer and a lower layer in the double-layer license plate, so that when license plate characters of the second image are segmented, the segmentation result can be more accurate.

The first obtaining sub-module is configured to segment the second image according to the vertical projection drawing to obtain N character images to be corrected included in the second image;. Specifically, the first determining sub-module may be configured to:. The second obtaining submodule is configured to segment the second image according to the vertical projection drawing to obtain N character images to be corrected included in the second image;.

The modules to of the apparatus are the same as the modules to of the embodiment shown in FIG. A fourth determining sub-module , configured to determine, before obtaining, according to a vertical projection method, a vertical projection diagram of a pixel value of each second image, according to a width-to-height ratio of the second image and a preset ratio threshold, whether the second image belongs to a double-layer license plate;. And the image character recognition module is configured to recognize characters in the N character images according to pre-stored character features of each country and region, and obtain a license plate number corresponding to the first image.

Wherein the module may be identical to the module in fig. Since the device embodiment is obtained based on the method embodiment and has the same technical effect as the method, the technical effect of the device embodiment is not described herein again.

For the apparatus embodiment, since it is substantially similar to the method embodiment, it is described relatively simply, and reference may be made to some descriptions of the method embodiment for relevant points. It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.

Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element. It will be understood by those skilled in the art that all or part of the steps in the above embodiments can be implemented by hardware associated with program instructions, and the program can be stored in a computer readable storage medium.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

A license plate number recognition method is characterized by comprising the following steps:. The method according to claim 1, wherein the preset condition is one or more of the following conditions:.