In real-time video streaming applications based on battery-powered camera systems, it is crucial to capture and transmit critical information satisfying the storage capacity and channel bandwidth requirement. In this paper, we propose a single-pass content-aware video encoding scheme for such applications to achieve a consistent video quality while satisfying the bit-rate constraints. To achieve the objective we developed: 1) scene descriptor based on the discrete cosine transformation coefficients obtained from the video encoder; 2) a parameterized analytic distortion model for video encoding using the scene descriptor; 3) scene change detection based on the encoded MB mode ratio; and 4) a single-pass consistent video quality control scheme which exploits 1), 2), and 3). Experimental results show that the proposed method results in less deviation from the target distortion and less distortion variance than the up-to-date best result by up to 65.25% and 75.88%, respectively, while satisfying the given rate constraints. The computing time overhead in the proposed method was 0.560% and 0.571% for the variable and constrained variable bit-rate mode, respectively, which is negligible in nearly all embedded applications.
Read full abstract