Research On Rate Control Algorithm Of UAV Aerial Video

Unmanned Aerial Vehicle(UAV)is widely used in military,civilian and other fields.It can carry reconnaissance equipment to collect video information for humans,which brings a lot of convenience to human life.However,the amount of data of UAV aerial video is very large,and it must be encoded in the actual application.High Efficiency Video Coding(HEVC)is one of the most widely used video coding standards.Compared with H.264/AVC,HEVC can double the compression ratio under the condition that the video quality is basically unchanged.In the HEVC encoding process,in order to be able to allocate bits more reasonably and meet the limitations of UAV storage or transmission bandwidth,rate control technology is required.However,due to the particularity of the scene taken by UAV,the existing rate control algorithms cannot be well applied in the UAV aerial photography field.In this thesis,the rate control technology in HEVC is deeply studied.In view of the shortcomings of the existing rate control algorithm,the related technology has been improved to make it more suitable for the application of UAV aerial photography.The main work and achievements in this thesis are as follows:(1)The research status of video coding standards is fully investigated,and the overall framework and key technologies of the HEVC coding standard are studied in depth.Secondly,the basic principles of rate control technology are introduced,and the rate control proposals in various coding standards are analyzed and summarized,and the focus is on the JCTVC-K0103 rate control proposal in HEVC.(2)Aiming at the characteristics of UAV aerial video,in order to ensure that the coding buffer does not "overflow" and "underflow",and improve the coding quality of salient regions,a rate control algorithm based on the characteristics of aerial video is proposed.Firstly,a bottom-up method is used to extract salient regions from the current video,and salient weights of each frame and each Coding Tree Unit(CTU)are calculated.Then the saliency weight of each frame and the buffer status are used to guide the frame layer bit allocation,and the saliency weight of each CTU is used to guide the CTU layer bit allocation.Experiment result shows that compared with the rate control algorithm in the HM16.9 standard model,the proposed algorithm can maintain the buffer in a stable state,and improve the peak signal-to-noise ratio of the salient area by 0.52 d B on average while ensuring the overall quality of the video.(3)Aiming at the problem of encoding quality degradation caused by scene changes during UAV aerial photography,corresponding improved algorithms are proposed.Firstly,the inter-frame variation factor of YUV was used to detect the scene change frame;In addition,in order to prevent external conditions such as illumination from affecting the detection results,inter-frame complexity variation factor was further used to detect scene change frames.After detecting the scene change frame,the coding structure is adjusted appropriately and the coding bits are redistributed without adding I frame.Experiment result shows that compared with the HM16.9 standard model,the proposed algorithm can improve the peak signal-to-noise ratio of video sequences by0.69 d B and reduce the coding error by 0.11%,and the improvement of coding quality is more obvious at the scene change frame.