A digital distribution material packaging system based on artificial intelligence

Abstract

The present invention relates to the field of video processing technology, and more specifically to a digital distribution material packaging system based on artificial intelligence. The system includes: an acquisition unit for acquiring video content to be released and classifying the video content; a storage unit for storing video content; a segmentation unit for segmenting the video content and acquiring video clips; a calculation unit for acquiring the first maximum complexity, the second maximum complexity and the third maximum complexity of the adjacent first video frame in the video clip; a compression unit for inputting each of the first video frames into an adjusted video compression model to acquire the first compressed video frame and resolution; a sending unit for sending a data packet to a receiving end; and a receiving unit for receiving and sending a data packet. The present invention solves the problem of video packaging, improves video transmission efficiency, and ensures video quality.

Description

Technical Field

[0001] The present invention relates to the field of video processing technology, and more specifically to an artificial intelligence-based digital distribution material packaging system. Background Art

[0002] With the continuous development of technologies such as artificial intelligence and big data, video packaging technology is also moving towards intelligence and automation. For example, by intelligently analyzing video content through deep learning algorithms, key frames and important information in the video can be more accurately identified, so that targeted optimization processing can be performed during the packaging process to improve packaging and transmission efficiency and video quality. Similar prior art includes a Chinese patent with publication number CN111565331B, which proposes an optimization method for wireless transmission of video image data. The optimization method for wireless transmission of video image data involves dividing the video into several video segments through a first video processing module, encoding each segment to generate a coding table, performing hash value detection on each video segment to generate several second hash values, and then transmitting the generated video segments to a second video processing module, and then detecting the transmitted video segments through the second video processing module, and verifying the integrity of the video segments in combination with the second hash value and the coding table, which can improve both the transmission efficiency and the integrity of video transmission, and realize the optimized transmission of video image data, overcoming the defects of slow transmission speed and poor transmission quality in the prior art. In addition, similar prior art includes a U.S. patent with publication number US8214857B2, which provides a method, system and product for generating a combined video stream from multiple input video streams. The processing includes generating a first video stream for a first product advertisement played at a first playback speed and a second video stream for a second product advertisement played at a second playback speed. The second playback speed is faster than the first playback speed. A combined video stream is generated, which includes a first panel in each frame to display the first video stream, and includes a second panel in each frame to display the second video stream. When the combined video stream is played at the first playback speed, the second video stream will appear slower than the second playback speed for generating the second video stream. Both of the above patents solve the method of video packaging, but neither of them considers dynamically obtaining compressed frames based on video frames and dynamically adjusting the sending method of video clips based on the communication bandwidth. Summary of the invention

[0003] In order to better solve the above problems, the present invention provides a digital distribution material packaging system based on artificial intelligence, the system comprising:

[0004] An acquisition unit, used for acquiring video content to be released, classifying the video content, and storing the video content in a storage unit corresponding to the classification;

[0005] A storage unit, used for storing the video content;

[0006] A segmentation unit, used to segment the video content and obtain video clips;

[0007] a calculation unit, configured to obtain a first maximum complexity, a second maximum complexity of a first video frame in the video segment, and a third maximum complexity of an adjacent first video frame;

[0008] a compression unit, configured to adjust the weight of a trained video compression model according to the first maximum complexity, the second maximum complexity, and the third maximum complexity, and input each of the first video frames into the adjusted video compression model to obtain a first compressed video frame and a resolution of the first compressed video frame;

[0009] The sending unit acquires a second packet and a sending data packet corresponding to the video clip to be sent according to the first bandwidth, the second bandwidth and the third bandwidth, wherein the first bandwidth is the actual communication bandwidth from the sending unit to the receiving unit, the second bandwidth is the occupied bandwidth required when the encoded first packet is used as the sending data packet, and the third bandwidth is the bandwidth occupied when the encoded third packet is used as the sending data packet;

[0010] The receiving unit is used to receive the sent data packet.

[0011] As a preferred technical solution, the sending unit is further configured as:

[0012] When the second bandwidth is less than the first bandwidth, the third bandwidth occupied by the third group after encoding is sent is calculated; when the third bandwidth is less than or equal to the first bandwidth, the third group after encoding is used as the sending data packet; when the third bandwidth is greater than the first bandwidth or the second bandwidth is equal to the first bandwidth, the first group after encoding is used as the sending data packet, wherein the first group is all the first compressed video frames corresponding to the video clip to be sent, and the third group is all the first video frames corresponding to the video clip to be sent; when the second bandwidth is greater than the first bandwidth, N subgroups are obtained based on all the first compressed videos corresponding to the video clip to be sent, and the first compressed video frames in the N subgroups are secondary compressed, and the N subgroups after the secondary compression and the remaining other first compressed video frames are used as the second group, and the encoded second group is sent to the receiving unit as a sending data packet.

[0013] As a preferred technical solution, the segmentation unit is used to compare adjacent video frames in the video content, obtain a comparison result, and take multiple adjacent video frames with the same scene as the same video segment;

[0014] The calculation unit is configured to: obtain each of the first video frames in the video clip, calculate the first pixel value difference and the second pixel difference between each pixel in each of the first video frames and other adjacent pixels in the horizontal direction and the vertical direction, and use the sum of the first pixel value difference and the sum of the second pixel difference corresponding to all pixels in the first video frame as the first complexity and the second complexity of the first video frame, respectively, calculate the third pixel value difference of pixels at the same position in adjacent first video frames, and use the sum of the third pixel value difference corresponding to all pixels in adjacent first video frames as the third complexity, and also use the maximum value corresponding to the first complexity, the maximum value of the second complexity and the maximum value of the third complexity in the first video frame as the first maximum complexity, the second maximum complexity and the third maximum complexity, respectively.

[0015] As a preferred technical solution, the compression unit is also used to normalize the first maximum complexity, the second maximum complexity and the third complexity, and multiply the first value corresponding to the normalized first maximum complexity, the second value corresponding to the second maximum complexity and the third value corresponding to the third maximum complexity with the first weight of the edge related to horizontal compression, the second weight of the edge related to vertical compression and the third weight of the edge related to compression resolution in each network layer of the video compression model, and use the result of the multiplication as the new first weight, the second weight and the third weight, obtain the maximum value of the first value, the second value and the third value, calculate the difference between the set value and the maximum value, multiply the difference with other weights of other edges in each network layer to obtain new other weights, wherein the set value is 1;

[0016] Each of the first video frames corresponding to the video clip is input into the adjusted video compression model to obtain a plurality of the first compressed video frames and the resolution of the first compressed video frames.

[0017] As a preferred technical solution, the computing unit is configured as follows:

[0018] acquiring a first data volume received and processed by the receiving unit per unit time, taking the first data volume as the first bandwidth, and periodically sending the first data volume to the sending unit through a feedback signal,

[0019] The ratio of the second data volume of the first group corresponding to the video segment to be sent after being encoded to the playback time of the video segment before the video segment to be sent is used as the second bandwidth;

[0020] The ratio of the third data volume of the third group corresponding to the video segment to be sent after being encoded to the playing time of the video segment before the video segment to be sent is used as the third bandwidth.

[0021] As an optimal technical solution, the compression unit is configured to: when the second bandwidth is greater than the first bandwidth, obtain m consecutive first compressed video frames with the greatest similarity and including the same subtitles from all first compressed video frames corresponding to the video clip to be sent, and use K groups of m first compressed video frames as K subgroups, and select N subgroups from the K subgroups, perform secondary compression on each subgroup, including using the first compressed video frame that is most forward in time in the subgroup as a reference video frame, and using other video frames after the reference video frame as difference video frames, and using the N subgroups corresponding to the video clip to be sent after secondary compression and the remaining other first compressed video frames as a second group, wherein the difference video frame is a difference video frame between the first compressed video frame and the previous first compressed video frame, wherein the value of m is a positive integer greater than or equal to 1, and the difference in the number of the first compressed video frames in any two subgroups is less than or equal to a set threshold.

[0022] As a preferred technical solution, the value of the subgroup N includes:

[0023] Calculate the bandwidth difference between the second bandwidth and the first bandwidth, sort the K subgroups from large to small according to the compression rate, and select N subgroups according to the sorting and the bandwidth difference, where the value of K is greater than or equal to N, and N is the minimum number of subgroups that satisfies that the second bandwidth corresponding to the second group is less than or equal to the first bandwidth.

[0024] As a preferred technical solution, the system further includes:

[0025] A decoding unit is used to decode each of the first compressed video frames by a decoding algorithm corresponding to the encoding algorithm when the receiving unit receives the sending data packet corresponding to the first group.

[0026] As a preferred technical solution, the encoding unit is also used to, when receiving the sending data packet corresponding to the second group, first decode the first compressed video frame in the second group and each secondary encrypted sub-group by a decoding algorithm corresponding to the encoding algorithm, and restore each of the first compressed video frames in the sub-group one by one by using the reference video frame and the difference video frame in the secondary encrypted sub-group.

[0027] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0028] The present invention divides the above-mentioned video content into the above-mentioned multiple video clips according to the scene through the above-mentioned segmentation unit, and also calculates the first complexity and the second complexity of each first video frame in the horizontal direction and the vertical direction respectively through the above-mentioned calculation unit, and also obtains the third complexity between adjacent frames, and obtains the first maximum complexity, the second maximum complexity and the third maximum complexity corresponding to the first video frame in the above-mentioned video clip, so as to facilitate the unified compression of the first video frames in the above-mentioned video clip. In order to ensure that the first video frame with the largest complexity within the frame and between adjacent frames has a higher picture quality, the above-mentioned first maximum complexity, the second maximum complexity and the third maximum complexity are used as the basis for obtaining the compression parameters of the above-mentioned video clip corresponding to the above-mentioned first video frame, and then different video clips can be obtained. By normalizing the above-mentioned first maximum complexity, the second maximum complexity and the above-mentioned third maximum complexity, it is convenient to adjust the weight of the video compression model, so that a video compression model suitable for the pictures of all the first video frames in the above-mentioned video clip can be obtained, and the video after adjustment The compression model obtains compression parameters suitable for each of the first video frames in the video clip and a first compressed video with no loss in picture quality, thereby not only reducing the transmission data but also ensuring the video quality. When the second bandwidth is less than the first bandwidth, that is, the actual communication bandwidth between the receiving end and the sending unit is sufficient to send the first packet, it is necessary to continue to determine whether the first bandwidth is sufficient to send the third packet. If it is sufficient, the encoded third packet is used as the third bandwidth. When the second bandwidth is equal to the first bandwidth or the third bandwidth is greater than the first bandwidth, the encoded first packet is used as the sending data packet. When the second bandwidth is greater than the first bandwidth, it is necessary to perform secondary compression on the corresponding N subpackets in the video clip to obtain the second packet, and the encoded second packet is used as the sending data packet. Through the mutual cooperation of the above technical solutions, not only the sending of the video clip can be guaranteed, but also the video picture quality can be guaranteed to be free of loss to the greatest extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a structural diagram of a digital distribution material packaging system based on artificial intelligence in the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0031] The present invention provides a digital distribution material packaging system based on artificial intelligence, such as Figure 1 As shown, the system comprises:

[0032] An acquisition unit, used for acquiring video content to be released, classifying the video content, and storing the video content in a storage unit corresponding to the classification;

[0033] A storage unit, used for storing the video content;

[0034] A segmentation unit, configured to read the video content from the storage unit, segment the video content, and obtain video segments;

[0035] a calculation unit, configured to obtain a first maximum complexity, a second maximum complexity of a first video frame in the video segment, and a third maximum complexity of an adjacent first video frame;

[0036] a compression unit, configured to adjust the weight of a trained video compression model according to the first maximum complexity, the second maximum complexity, and the third maximum complexity, and input each of the first video frames into the adjusted video compression model to obtain a first compressed video frame and a resolution of the first compressed video frame;

[0037] The sending unit acquires a second packet and a sending data packet corresponding to the to-be-sent video clip according to a first bandwidth, a second bandwidth, and a third bandwidth, wherein the first bandwidth is an actual communication bandwidth from the sending unit to a receiving end, the second bandwidth is an occupied bandwidth required when the encoded first packet is used as the sending data packet, and the third bandwidth is a bandwidth occupied when the encoded third packet is used as the sending data packet;

[0038] The receiving end is used to receive the data packet.

[0039] Specifically, the video content is acquired through the acquisition unit and saved to the storage unit, and the video content is read through the division unit. The video content is divided into the multiple video segments according to the scene through the segmentation unit, and the first complexity and the second complexity of each first video frame in the horizontal and vertical directions are calculated respectively through the calculation unit, and the third complexity between adjacent frames is acquired. The first maximum complexity, the second maximum complexity and the third maximum complexity corresponding to the first video frame in the video segment are acquired, so as to uniformly compress the first video frame in the video segment, in order to ensure the first video frame with the largest complexity within the frame and between adjacent frames. All of them have high picture quality, so the above-mentioned first maximum complexity, the second maximum complexity and the third maximum complexity are used as the basis for obtaining the compression parameters of the above-mentioned first video frame corresponding to the above-mentioned video clip, and then different video clips can be obtained. By normalizing the above-mentioned first maximum complexity, the second maximum complexity and the above-mentioned third maximum complexity, it is convenient to adjust the weight of the video compression model, so that a video compression model suitable for the pictures of all the first video frames in the above-mentioned video clip can be obtained, and by inputting each of the above-mentioned first video frames in the above-mentioned video clip into the adjusted above-mentioned video compression model, the first compressed video frame corresponding to the above-mentioned first video frame and the resolution of the above-mentioned first compressed video frame are obtained. , it is possible to obtain the compression parameters adapted to each of the first video frames in the video clip and the first compressed video without loss of picture quality, thereby not only reducing the transmission data but also ensuring the video quality. When the second bandwidth is less than the first bandwidth, that is, the actual communication bandwidth between the receiving end and the sending unit is sufficient to send the first packet, that is, all the first compressed videos corresponding to the encoded video clip to be sent, at this time, it is necessary to continue to determine whether the first bandwidth is sufficient to send the third packet, that is, the first video frame corresponding to the video clip to be sent. If it is sufficient, that is, when the third bandwidth is less than or equal to the first bandwidth, the encoded third packet is used as the The third bandwidth, when the second bandwidth is equal to the first bandwidth or the third bandwidth is greater than the first bandwidth, that is, the first bandwidth is just enough to send the first packet or is not enough to send the third packet, the encoded first packet is used as the sending data packet; when the second bandwidth is greater than the first bandwidth, that is, the first bandwidth is not enough to send the first packet, it is necessary to perform secondary compression on the corresponding N subpackets in the video clip to obtain the second packet, and use the encoded second packet as the sending data packet. Through the above technical solution, not only the sending of the video clip can be guaranteed, but also the video picture quality can be guaranteed to be not lost to the greatest extent.

[0040] Furthermore, the sending unit is further configured as:

[0041] When the second bandwidth is less than the first bandwidth, the third bandwidth occupied by the third group after encoding is sent is calculated; when the third bandwidth is less than or equal to the first bandwidth, the third group after encoding is used as the sending data packet; when the third bandwidth is greater than the first bandwidth or the second bandwidth is equal to the first bandwidth, the first group after encoding is used as the sending data packet, wherein the first group is all the first compressed video frames corresponding to the video clip to be sent, and the third group is all the first video frames corresponding to the video clip to be sent; when the second bandwidth is greater than the first bandwidth, N subgroups are obtained based on all the first compressed videos corresponding to the video clip to be sent, and the first compressed video frames in the N subgroups are secondary compressed, and the N subgroups after the secondary compression and the remaining other first compressed video frames are used as the second group, and the encoded second group is sent to the receiving unit as a sending data packet.

[0042] Specifically, when the second bandwidth is less than the first bandwidth, that is, the actual communication bandwidth between the receiving end and the sending unit is sufficient to send the first packet, that is, all the first compressed videos corresponding to the encoded video clip to be sent. At this time, it is necessary to continue to determine whether the first bandwidth is sufficient to send the third packet, that is, the first video frame corresponding to the video clip to be sent. If it is sufficient, that is, the third bandwidth is less than or equal to the first bandwidth, the encoded third packet is used as the third bandwidth. When the second bandwidth is equal to the first bandwidth or the third bandwidth is greater than the first bandwidth, that is, the first bandwidth is just sufficient for the first packet to be sent or is insufficient to send the third packet, the encoded first packet is used as the sending data packet. When the second bandwidth is greater than the first bandwidth, that is, the first bandwidth is insufficient to send the first packet, it is necessary to perform secondary compression on the corresponding N subpackets in the video clip to obtain the second packet, and the encoded second packet is used as the sending data packet. Through the above technical solution, not only the sending of the video clip can be guaranteed, but also the video picture quality can be guaranteed to be not lost to the greatest extent.

[0043] Furthermore, the segmentation unit is used to compare adjacent video frames in the video content, obtain a comparison result, and take the multiple adjacent video frames of the same scene in the comparison result as the same video segment;

[0044] The calculation unit is configured to: obtain each of the first video frames in the video clip, calculate the first pixel value difference and the second pixel difference between each pixel in each of the first video frames and other adjacent pixels in the horizontal direction and the vertical direction, and use the sum of the first pixel value difference and the sum of the second pixel difference corresponding to all pixels in the first video frame as the first complexity and the second complexity of the first video frame, respectively, calculate the third pixel value difference of pixels at the same position in adjacent first video frames, and use the sum of the third pixel value difference corresponding to all pixels in adjacent first video frames as the third complexity, and also use the maximum value of the first complexity, the maximum value of the second complexity and the maximum value of the third complexity in the first video frame as the first maximum complexity, the second maximum complexity and the third maximum complexity, respectively.

[0045] Specifically, the video content is divided into the above-mentioned multiple video segments according to the scene by the above-mentioned segmentation unit. Since the image difference of the first video frame in the video segment of the same scene is small, it is convenient to use the same compression parameters for unified compression processing without losing the picture quality. The first complexity and the second complexity of each first video frame in the horizontal direction and the vertical direction are calculated by the above-mentioned calculation unit, wherein the first complexity and the second complexity reflect the complexity within the frame, and the third complexity between adjacent frames is obtained, and the first maximum complexity, the second maximum complexity and the third maximum complexity corresponding to the first video frame in the above-mentioned video segment are obtained. Since the first video frame in the above-mentioned video segment needs to be uniformly compressed, in order to ensure that the first video frame with the largest complexity within the frame and between adjacent frames has a higher picture quality, the first maximum complexity, the second maximum complexity and the third maximum complexity are used as the basis for obtaining the compression parameters of the above-mentioned video segment corresponding to the above-mentioned first video frame, and then different video segments can be obtained, thereby ensuring that the first video frame in the above-mentioned video segment reduces the transmission pressure after compression, and can also ensure that the restored picture quality is not lost.

[0046] Further, the compression unit is also used to normalize the first maximum complexity, the second maximum complexity and the third complexity, and multiply the first value corresponding to the normalized first maximum complexity, the second value corresponding to the second maximum complexity and the third value corresponding to the third maximum complexity with the first weight of the edge related to horizontal compression, the second weight of the edge related to vertical compression and the third weight of the edge related to compression resolution in each network layer of the video compression model, and use the result of the multiplication as the new first weight, the second weight and the third weight, obtain the maximum value of the first value, the second value and the third value, calculate the difference between the set value and the maximum value, multiply the difference with other weights of other edges in each network layer to obtain new other weights, wherein the set value is 1;

[0047] Each of the first video frames corresponding to the video clip is input into the adjusted video compression model to obtain the corresponding first compressed video frame and the resolution of the first compressed video frame.

[0048] Specifically, by normalizing the first maximum complexity, the second maximum complexity and the third maximum complexity, the values ​​of the three are converted into [0,1], so as to facilitate weight adjustment of the video compression model, the product of the first value and the first weight is used as the new first weight, the product of the second value and the second weight is used as the new second weight, the product of the third value and the third weight is used as the new third weight, and the difference between the set value and the maximum value of the first value, the second value and the third value is obtained, and the difference is multiplied by other weights of other edges in each network layer to obtain new other weights, so as to obtain a video compression model suitable for the pictures of all the first video frames in the video clip, and by inputting each of the first video frames in the video clip into the adjusted video compression model, the first compressed video frame corresponding to the first video frame and the resolution of the first compressed video frame are obtained. Through the above technical solution, the compression parameters suitable for each of the first video frames in the video clip and the first compressed video with no loss of picture quality can be obtained, thereby not only reducing the transmission data but also ensuring the video quality.

[0049] Furthermore, the computing unit is configured as follows:

[0050] acquiring a first data volume received and processed by the receiving unit per unit time, taking the first data volume as the first bandwidth, and periodically sending the first data volume to the sending unit through a feedback signal,

[0051] The ratio of the second data volume of the first group corresponding to the video segment to be sent after being encoded to the playback time of the video segment before the video segment to be sent is used as the second bandwidth;

[0052] The ratio of the third data volume of the third group corresponding to the video segment to be sent after being encoded to the playing time of the video segment before the video segment to be sent is used as the third bandwidth.

[0053] Specifically, the first bandwidth reflects the data transmission capacity of the communication link between the sending unit and the receiving end. When the receiving end is a user end, the worst case is to play the video content while receiving it. Therefore, the playing time of the previous video segment of the video segment to be sent is used as the shortest sending time, the second data amount of the first group corresponding to the video to be sent is used as the maximum sending data amount, and the ratio of the maximum sending data amount to the shortest sending time is used as the second bandwidth. Through the technical solution, a foundation is laid for adjusting the sending method of the video segment to be sent according to the first bandwidth and the second bandwidth.

[0054] Further, the compression unit is configured to: when the second bandwidth is greater than the first bandwidth, obtain m consecutive first compressed video frames with the greatest similarity and including the same subtitles from all the first compressed video frames corresponding to the video clip to be sent, and use K groups of m first compressed video frames as K subgroups, and select N subgroups from the K subgroups, and perform secondary compression on each subgroup, including using the first compressed video frame that is most forward in time in the subgroup as a reference video frame, and using other video frames after the reference video frame as difference video frames, and using the N subgroups corresponding to the video clip to be sent after secondary compression and the remaining other first compressed video frames as a second group, wherein the difference video frame is a difference video frame between the first compressed video frame and the previous first compressed video frame, wherein the value of m is a positive integer greater than or equal to 1, and the difference in the number of the first compressed video frames in any two subgroups is less than or equal to a set threshold.

[0055] Specifically, when the second bandwidth is greater than the first bandwidth, that is, when the actual communication bandwidth between the sending unit and the receiving end is less than the occupied bandwidth of the sending data packet, it is necessary to further compress the first compressed video frame corresponding to the video segment, so that the sending data packet corresponding to the to-be-sent video segment can be sent within the waiting time. By obtaining the continuous m first compressed video frames with the greatest similarity and including the same subtitle, i.e., the same subtitle content, from all the first compressed video frames corresponding to the video segment, and taking the continuous m first compressed video frames as a subgroup, K subgroups are obtained by the same method, and the remaining other first compressed video frames from the N subgroups and all the first compressed video frames corresponding to the to-be-sent video segment are taken as the second group. Since the first compressed video frames in the subgroups are all video frames with high similarity, the first compressed video frame with the earliest corresponding playback time in the subgroup is taken as the reference video frame, and the difference video frames of each first compressed video frame after the reference video frame relative to the previous first compressed video frame are obtained in sequence, that is, only different pixels are retained and the same pixels are eliminated, thereby achieving secondary compression, saving the sending bandwidth, and not affecting the video picture quality.

[0056] Furthermore, the value of the number N of subgroups includes:

[0057] Calculate the bandwidth difference between the second bandwidth and the first bandwidth, sort the K subgroups from large to small according to the compression rate, and select N subgroups according to the sorting and the bandwidth difference, where the value of K is greater than or equal to N, and N is the minimum number of subgroups that satisfies that the second bandwidth corresponding to the second group is less than or equal to the first bandwidth.

[0058] Specifically, since the above-mentioned first bandwidth, that is, the actual communication bandwidth may change with the amount of communication data of the receiving end or the sending unit, in order to adapt to this dynamic change, when the above-mentioned second bandwidth is greater than the above-mentioned first bandwidth, the K above-mentioned sub-packets are obtained and sorted according to the compression rate. The larger the difference in the video frame size before and after the compression of the above-mentioned sub-packets is, the greater the above-mentioned compression rate is. In addition, by calculating the bandwidth difference between the above-mentioned second bandwidth and the above-mentioned first bandwidth, and selecting N above-mentioned sub-packets based on the above-mentioned bandwidth difference and the above-mentioned sorting, N is the minimum number of sub-packets that satisfies the above-mentioned second group corresponding to the second bandwidth being less than or equal to the first bandwidth. Through the above-mentioned technical solution, when the above-mentioned second bandwidth is greater than the above-mentioned first bandwidth, the number of the above-mentioned sub-packets is dynamically adjusted, which can not only ensure the smooth sending of the sending data packet corresponding to the video clip to be sent, but also maximize the quality of the video frame.

[0059] Furthermore, the system further comprises:

[0060] A decoding unit is used to decode each of the first compressed video frames and each of the first video frames respectively by using a decoding algorithm corresponding to the encoding algorithm when the receiving end receives the transmitted data packets corresponding to the encoded first group and the third group.

[0061] Specifically, through the above technical solution, the first compressed video frame and the first video frame corresponding to the above first group and the above third group respectively can be accurately decoded.

[0062] Furthermore, the encoding unit is also used to, when receiving the sending data packet corresponding to the second group, first decode the first compressed video frame in the second group and each secondary encrypted sub-group by a decoding algorithm corresponding to the encoding algorithm, and restore each first compressed video frame in the sub-group one by one by using the reference video frame and the difference video frame in the secondary encrypted sub-group.

[0063] Specifically, through the above technical solution, all the above first compressed video frames corresponding to the above second grouping can be accurately decoded.

[0064] In summary, the present invention divides the above-mentioned video content into the above-mentioned multiple video segments according to the scene through the above-mentioned segmentation unit, and also calculates the first complexity and the second complexity of each first video frame in the horizontal direction and the vertical direction through the above-mentioned calculation unit, and also obtains the third complexity between adjacent frames, and obtains the first maximum complexity, the second maximum complexity and the third maximum complexity corresponding to the first video frame in the above-mentioned video segment, so as to facilitate the unified compression of the first video frames in the above-mentioned video segment. In order to ensure that the first video frame with the largest complexity within the frame and between adjacent frames has a higher picture quality, the above-mentioned first maximum complexity, the second maximum complexity and the third maximum complexity are used as the basis for obtaining the compression parameters of the above-mentioned video segment corresponding to the above-mentioned first video frame, and then different video segments can be obtained. By normalizing the above-mentioned first maximum complexity, the second maximum complexity and the above-mentioned third maximum complexity, it is convenient to adjust the weight of the video compression model, so that a video compression model suitable for the pictures of all the first video frames in the above-mentioned video segment can be obtained. After adjustment The video compression model obtains the compression parameters and the first compressed video without loss of picture quality suitable for each of the first video frames in the video clip, thereby not only reducing the transmission data but also ensuring the video quality. When the second bandwidth is less than the first bandwidth, that is, the actual communication bandwidth between the receiving end and the transmitting end is sufficient to send the first packet, at this time, it is necessary to continue to judge whether the first bandwidth is sufficient to send the third packet. If it is sufficient, the encoded third packet is used as the third bandwidth. When the second bandwidth is equal to the first bandwidth or the third bandwidth is greater than the first bandwidth, the encoded first packet is used as the sending data packet. When the second bandwidth is greater than the first bandwidth, it is necessary to perform secondary compression on the corresponding N subpackets in the video clip to obtain the second packet, and the encoded second packet is used as the sending data packet. Through the mutual cooperation of the above technical solutions, not only the sending of the above video clip can be guaranteed, but also the video picture quality can be guaranteed to be not lost to the greatest extent.

[0065] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0066] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0067] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A digital distribution material packaging system based on artificial intelligence, characterized in that: The system comprises: An acquisition unit, used for acquiring video content to be released, classifying the video content, and storing the video content in a storage unit corresponding to the classification; A storage unit, used for storing the video content; A segmentation unit, used to segment the video content and obtain video clips; a calculation unit, configured to obtain a first maximum complexity, a second maximum complexity of a first video frame in the video segment, and a third maximum complexity of an adjacent first video frame; a compression unit, configured to adjust the weight of a trained video compression model according to the first maximum complexity, the second maximum complexity, and the third maximum complexity, and input each of the first video frames into the adjusted video compression model to obtain a first compressed video frame and a resolution of the first compressed video frame; A sending unit, which acquires a second packet and a sending data packet corresponding to the to-be-sent video clip according to a first bandwidth, a second bandwidth, and a third bandwidth, wherein the first bandwidth is an actual communication bandwidth from the sending unit to the receiving unit, the second bandwidth is a bandwidth occupied when the encoded first packet is used as the sending data packet, and the third bandwidth is a bandwidth occupied when the encoded third packet is used as the sending data packet; The receiving unit is used to receive the sent data packet; The sending unit is further configured as: When the second bandwidth is less than the first bandwidth, the third bandwidth occupied by the third group after encoding is sent is calculated; when the third bandwidth is less than or equal to the first bandwidth, the third group after encoding is used as the sending data packet; when the third bandwidth is greater than the first bandwidth or the second bandwidth is equal to the first bandwidth, the first group after encoding is used as the sending data packet, wherein the first group is all the first compressed video frames corresponding to the video clip to be sent, and the third group is all the first video frames corresponding to the video clip to be sent; when the second bandwidth is greater than the first bandwidth, N subgroups are obtained based on all the first compressed videos corresponding to the video clip to be sent, and the first compressed video frames in the N subgroups are secondary compressed, and the N subgroups after secondary compression and the remaining other first compressed video frames are used as the second group, and the second encoded group is sent to the receiving unit as a sending data packet; The segmentation unit is used to compare adjacent video frames in the video content, obtain a comparison result, and take multiple adjacent video frames with the same scene as the same video segment; The calculation unit is configured to: obtain each of the first video frames in the video clip, calculate the first pixel value difference and the second pixel difference between each pixel in each of the first video frames and other adjacent pixels in the horizontal direction and the vertical direction, and use the sum of the first pixel value difference and the sum of the second pixel difference corresponding to all pixels in the first video frame as the first complexity and the second complexity of the first video frame, respectively, calculate the third pixel value difference between the first video frame and each pixel in the same position of the adjacent first video frames, and use the sum of the third pixel value differences corresponding to all pixels in the adjacent first video frames as the third complexity, and also use the maximum value of the first complexity, the maximum value of the second complexity and the maximum value of the third complexity corresponding to all the first video frames as the first maximum complexity, the second maximum complexity and the third maximum complexity, respectively.

2. The system according to claim 1, characterized in that The compression unit is further used to normalize the first maximum complexity, the second maximum complexity and the third maximum complexity, and multiply the first value corresponding to the normalized first maximum complexity, the second value corresponding to the second maximum complexity and the third value corresponding to the third maximum complexity by the first weight of the edge related to horizontal compression, the second weight of the edge related to vertical compression and the third weight of the edge related to compression resolution in each network layer of the video compression model, and use the result of the multiplication as the new first weight, the second weight and the third weight, obtain the maximum value of the first value, the second value and the third value, calculate the difference between the set value and the maximum value, and multiply the difference by other weights of other edges in each network layer to obtain the new other weights, wherein the set value is 1; Each of the first video frames corresponding to the video clip is input into the adjusted video compression model to obtain a plurality of the first compressed video frames and the resolution of the first compressed video frames.

3. The system according to claim 1, characterized in that The computing unit is configured as follows: acquiring a first data volume received and processed by the receiving unit per unit time, taking the first data volume as the first bandwidth, and periodically sending the first data volume to the sending unit through a feedback signal, The ratio of the second data volume of the first group corresponding to the video segment to be sent after being encoded to the playback time of the video segment before the video segment to be sent is used as the second bandwidth; The ratio of the third data volume of the third group corresponding to the video segment to be sent after being encoded to the playing time of the video segment before the video segment to be sent is used as the third bandwidth.

4. The system according to claim 1, characterized in that The compression unit is configured to: when the second bandwidth is greater than the first bandwidth, obtain m consecutive first compressed video frames with the greatest similarity and including the same subtitles from all first compressed video frames corresponding to the video clip to be sent, and use K groups of m first compressed video frames as K subgroups, select N subgroups from the K subgroups, and perform secondary compression on each subgroup, including using the first compressed video frame that is the earliest in time in the subgroup as a reference video frame, and using other video frames after the reference video frame as difference video frames, and using the N subgroups corresponding to the video clip to be sent after secondary compression and the remaining other first compressed video frames as a second group, wherein the difference video frame is a difference video frame between the first compressed video frame and the previous first compressed video frame, wherein the value of m is a positive integer greater than or equal to 1, and the difference in the number of the first compressed video frames in any two subgroups is less than or equal to a set threshold.

5. The system according to claim 4, characterized in that The value of the number N of subgroups includes: Calculate the bandwidth difference between the second bandwidth and the first bandwidth, sort the K subgroups from large to small according to the compression rate, and select N subgroups according to the sorting and the bandwidth difference, where the value of K is greater than or equal to N, and N is the minimum number of subgroups that satisfies that the second bandwidth corresponding to the second group is less than or equal to the first bandwidth.

6. The system according to claim 1, characterized in that The system further comprises: A decoding unit is used to decode each of the first compressed video frames by a decoding algorithm corresponding to the encoding algorithm when the receiving unit receives the sending data packet corresponding to the first group.

7. The system according to claim 4, characterized in that The encoding unit is also used to, when receiving the sending data packet corresponding to the second group, first decode the first compressed video frame in the second group and each secondary encrypted sub-group by a decoding algorithm corresponding to the encoding algorithm, and restore each of the first compressed video frames in the sub-group one by one by using the reference video frame and the difference video frame in the secondary encrypted sub-group.

Images