阅读说明：本技术 参考数据实时上传机构 (Reference data real-time uploading mechanism ) 是由 不公告发明人 于 2019-05-30 设计创作，主要内容包括：本发明涉及一种参考数据实时上传机构,所述机构包括：数据解析设备,用于基于女性人体成像特征和男性人体成像特征分别从实时滤波图像中提取各个女性人体对象和各个男性人体对象,并基于各个女性人体对象和各个男性人体对象计算当前女性观看比例和当前男性观看比例；内容上传设备,用于在所述电视播放终端当前播放内容为广告时,触发当前女性观看比例和当前男性观看比例的上传。本发明的参考数据实时上传机构运行可靠、数据有效。由于在电视播放终端当前播放内容为广告时,将观看电视的男女比例和广告标识一起打包发送给远端的广告服务商,从而为广告服务商广告的针对性发布提供有价值的参考数据。(The invention relates to a reference data real-time uploading mechanism, which comprises: the data analysis equipment is used for respectively extracting each female human body object and each male human body object from the real-time filtering image based on the female human body imaging characteristics and the male human body imaging characteristics, and calculating the current female viewing proportion and the current male viewing proportion based on each female human body object and each male human body object; and the content uploading equipment is used for triggering the uploading of the current female watching proportion and the current male watching proportion when the current playing content of the television playing terminal is an advertisement. The reference data real-time uploading mechanism provided by the invention is reliable in operation and effective in data. When the current playing content of the television playing terminal is the advertisement, the male and female proportion of watching the television and the advertisement identification are packaged together and sent to the remote advertisement service provider, so that valuable reference data is provided for the targeted release of the advertisement service provider.)

1. The utility model provides a mechanism is uploaded in real time to reference data which characterized in that includes:

the data analysis equipment is connected with the image merging equipment and used for respectively extracting each female human body object and each male human body object from the real-time filtering image based on the female human body imaging characteristics and the male human body imaging characteristics and calculating the current female viewing proportion and the current male viewing proportion based on each female human body object and each male human body object;

the content uploading device is connected with the data analysis device and the set top box of the television playing terminal and is used for triggering the uploading of the current female watching proportion and the current male watching proportion when the current playing content of the television playing terminal is an advertisement;

triggering the uploading of the current female viewing proportion and the current male viewing proportion comprises: packaging the advertisement identification played currently, the current female watching proportion and the current male watching proportion and sending the packaged advertisement identification, the current female watching proportion and the current male watching proportion to a remote advertisement service provider;

the set-top box is an IP television set-top box, and the content uploading device packs the currently played advertisement identification, the current female watching proportion and the current male watching proportion through a network transmission interface of the IP television set-top box and sends the current advertisement identification, the current female watching proportion and the current male watching proportion to a remote advertisement service provider;

the monitoring camera is embedded in the shell of the television playing terminal and is used for carrying out camera shooting operation on a scene in front of the television playing terminal so as to obtain a corresponding front monitoring image;

the Lanczos interpolation equipment is connected with the monitoring camera and used for receiving the front monitoring image and executing Lanczos interpolation processing on the front monitoring image so as to obtain and output a corresponding Lanczos interpolation image;

the statistics sorting filtering equipment is connected with the Lanczos interpolation equipment and is used for performing statistics sorting filtering processing on the received Lanczos interpolation image so as to obtain and output a corresponding statistics sorting filtering image;

the pixel identification device is used for receiving the Lanczos interpolation image, carrying out binarization processing on the Lanczos interpolation image to obtain and output a corresponding binarized image, traversing each pixel of the image from the upper left corner of the binarized image, and regarding each pixel as an internal pixel if the pixels of more than four points around the pixel are all black-level pixels or the pixels of more than four points around the pixel are all white-level pixels, otherwise, regarding the pixel as a preliminary contour pixel;

the pixel identification equipment is also used for connecting all the preliminary contour pixels in the binary image and fitting the preliminary contour pixels to obtain a plurality of fitting closed curves, modifying the preliminary contour pixels outside the fitting closed curves into internal pixels, and taking the unmodified preliminary contour pixels as final contour pixels;

the command extraction device is connected with the pixel identification device and used for connecting all final contour pixels to obtain a plurality of closed regions, determining the area of each closed region, and sending a human body identification command when the closed region with the area equal to the preset human body area distribution range exists;

the command extraction equipment is also used for sending a human body unidentified command when no closed region with the area equal to the preset human body area distribution range exists;

the gradient sharpening filtering equipment is respectively connected with the Lanczos interpolation equipment and the command extraction equipment and is used for executing gradient sharpening filtering processing on the Lanczos interpolation image when the human body unidentified command is received so as to obtain and output a corresponding gradient sharpening filtering processing image;

the gradient sharpening filtering equipment is further used for not executing gradient sharpening filtering processing on the Lanczos interpolation image when the human body identification command is received;

and the image merging device is respectively connected with the statistics sorting filtering device and the gradient sharpening filtering device, and is used for outputting the statistics sorting filtering image as a real-time filtering image when the human body identification command is received, and is also used for outputting the gradient sharpening filtering processing image as a real-time filtering image when the human body mismatching command is received.

2. The real-time reference data uploading mechanism of claim 1, further comprising:

and the parameter identification equipment is connected with the monitoring camera and used for receiving the front monitoring image and identifying the frame rate of the image sequence where the front monitoring image is located so as to obtain and output a corresponding real-time frame rate.

3. The real-time reference data uploading mechanism of claim 2, further comprising:

and the signal sending equipment is connected with the parameter identification equipment, and is used for receiving the real-time frame rate and sending a first trigger command when the real-time frame rate exceeds a preset frame rate threshold value.

4. The real-time reference data uploading mechanism according to claim 3, wherein:

the signal sending device is further configured to send a second trigger command when the real-time frame rate is lower than or equal to the preset frame rate threshold.

5. The real-time reference data uploading mechanism according to claim 4, further comprising:

and the homomorphic filtering equipment is connected with the signal sending equipment and used for starting receiving the front monitoring image when receiving the first trigger command and executing homomorphic filtering processing on the front monitoring image so as to obtain a homomorphic filtering image.

6. The real-time reference data uploading mechanism according to claim 5, further comprising:

and the data identification equipment is connected with the homomorphic filtering equipment and used for receiving the homomorphic filtering image, adjusting the edge enhancement processing strength of the L component sub-image in the homomorphic filtering image LAB space based on a real-time frame rate, adjusting the edge enhancement processing strength of the A component sub-image in the homomorphic filtering image LAB space based on the real-time frame rate, and adjusting the edge enhancement processing strength of the B component sub-image in the homomorphic filtering image LAB space based on the real-time frame rate.

7. The real-time reference data uploading mechanism according to claim 6, further comprising:

and the self-adaptive processing equipment is connected with the data identification equipment and is used for parallelly executing edge enhancement processing with respective edge enhancement processing strength on the L-component sub-image, the A-component sub-image and the B-component sub-image in the homomorphic filtering image LAB space so as to obtain a corresponding self-adaptive processing image.

8. The real-time reference data uploading mechanism of claim 7, further comprising:

the histogram equalization equipment is respectively connected with the Lanczos interpolation equipment and the adaptive processing equipment and is used for receiving the adaptive processing image, executing histogram equalization processing on the adaptive processing image to obtain a histogram equalization image, and sending the histogram equalization image to the Lanczos interpolation equipment in place of the front monitoring image;

in the data identification device, the degree of change of the edge enhancement processing intensity of the L component sub-image along with the real-time frame rate is the same as the degree of change of the edge enhancement processing intensity of the A component sub-image along with the real-time frame rate;

in the data identification device, the degree of change of the edge enhancement processing intensity of the B component sub-image along with the real-time frame rate is lower than that of the edge enhancement processing intensity of the L component sub-image along with the real-time frame rate.

9. The real-time reference data uploading mechanism of claim 8, wherein:

in the data identification equipment, the real-time frame rate is in direct proportion to the edge enhancement processing strength of the L component sub-image in the homomorphic filtering image LAB space, the real-time frame rate is in direct proportion to the edge enhancement processing strength of the A component sub-image in the homomorphic filtering image LAB space, and the real-time frame rate is in direct proportion to the edge enhancement processing strength of the B component sub-image in the homomorphic filtering image LAB space;

wherein the homomorphic filtering device is further configured to stop receiving the front monitoring image upon receiving the second trigger command.