945 resultados para H.264
Resumo:
This paper discusses a novel high-speed approach for human action recognition in H.264/AVC compressed domain. The proposed algorithm utilizes cues from quantization parameters and motion vectors extracted from the compressed video sequence for feature extraction and further classification using Support Vector Machines (SVM). The ultimate goal of the proposed work is to portray a much faster algorithm than pixel domain counterparts, with comparable accuracy, utilizing only the sparse information from compressed video. Partial decoding rules out the complexity of full decoding, and minimizes computational load and memory usage, which can result in reduced hardware utilization and faster recognition results. The proposed approach can handle illumination changes, scale, and appearance variations, and is robust to outdoor as well as indoor testing scenarios. We have evaluated the performance of the proposed method on two benchmark action datasets and achieved more than 85 % accuracy. The proposed algorithm classifies actions with speed (> 2,000 fps) approximately 100 times faster than existing state-of-the-art pixel-domain algorithms.
Resumo:
In this paper, we have proposed an anomaly detection algorithm based on Histogram of Oriented Motion Vectors (HOMV) 1] in sparse representation framework. Usual behavior is learned at each location by sparsely representing the HOMVs over learnt normal feature bases obtained using an online dictionary learning algorithm. In the end, anomaly is detected based on the likelihood of the occurrence of sparse coefficients at that location. The proposed approach is found to be robust compared to existing methods as demonstrated in the experiments on UCSD Ped1 and UCSD Ped2 datasets.
Resumo:
Real time anomaly detection is the need of the hour for any security applications. In this article, we have proposed a real time anomaly detection for H.264 compressed video streams utilizing pre-encoded motion vectors (MVs). The proposed work is principally motivated by the observation that MVs have distinct characteristics during anomaly than usual. Our observation shows that H.264 MV magnitude and orientation contain relevant information which can be used to model the usual behavior (UB) effectively. This is subsequently extended to detect abnormality/anomaly based on the probability of occurrence of a behavior. The performance of the proposed algorithm was evaluated and bench-marked on UMN and Ped anomaly detection video datasets, with a detection rate of 70 frames per sec resulting in 90x and 250x speedup, along with on-par detection accuracy compared to the state-of-the-art algorithms.
Resumo:
A doença pulmonar obstrutiva crônica (DPOC) causa a redução da capacidade respiratória e seu desenvolvimento é associado à fumaça de cigarro. O cigarro possui mais de 4800 substâncias tóxicas e causa a morte de seis milhões de pessoas por ano no mundo. Estudos buscam meios de reverter os males causados pela fumaça de cigarro. A própolis (P) é um produto produzido por abelhas que possui várias propriedades. O objetivo deste trabalho foi avaliar os efeitos antioxidantes da P em macrófagos murinos e na inflamação pulmonar aguda induzida pela fumaça de cigarro (CS) em camundongos. A análise dos compostos fitoquímicos do extrato alcóolico da P (EAP) foi feita por cromatografia gasosa acoplada à espectrometria de massa (GC-MS). Células da linhagem RAW 264.7 foram tratadas em diversas concentrações de P durante 24 horas. Após tratamento, as seguintes análises foram realizadas: polifenóis totais; viabilidade celular (MTT); potencial redutor (DPPH); espécies reativas de oxigênio totais (ROS) e de malondialdeído (MDA). Trinta camundongos C57BL/6 foram divididos em 3 grupos (n=10/grupo): Controle+P, CS e CS+P. Ambos os grupos CS foram expostos a 6 cigarros/dia durante 5 dias. O grupo CS foi tratado com veículo. O pulmão e o lavado broncoalveolar (BAL) foram coletados para análise histológica e contagem diferencial de células. As análises para mieloperoxidase (MPO), superóxido dismutase (SOD), catalase (CAT), glutationa peroxidase (GPx), glutationa reduzida (GSH) e oxidada (GSSG), MDA, nitrito e western blotting para TNF-alfa foram realizadas. A análise fitoquímica do EAP mostrou a presença dos ácidos hidrocinâmicos e coumárico, a artepilina C e a drupanina. Foi observado o aumento concentração-dependente dos níveis de polifenóis totais (p<0,001), do MTT (p<0,001) e do DPPH (p<0,001), e o inverso com o MDA (p<0,001). Os níveis de ROS diminuem nas concentrações de 15,6 e 31,2 mg/mL (p<0,05, ambos). A histologia pulmonar do grupo Controle+P foi similar ao do CS+P e foi observado um influxo de macrófagos e neutrófilos no grupo CS (p<0,01 e p<0,001, respectivamente). Os níveis de MPO foram aumentados no grupo CS (526,534,72 mU/mg ptn, p<0,01), mas houve uma redução no grupo CS+P (385,127,64 mU/mg ptn, p<0,05) comparável ao Controle+P (13412,99 mU/mg ptn, p<0,001), o mesmo aconteceu com as enzimas antioxidantes: SOD (Controle+P: 523,529,6 U/mg ptn; CS: 523,529,6 U/mg ptn, p<0,001; CS+P: 246,815,69 U/mg ptn, p<0,001); CAT (Controle+P: 37,383,39 U/mg ptn; CS: 92,686,24 U/mg ptn, p<0,001; CS+P: 59,844,55 U/mg ptn, p<0,05); GPx (Controle+P: 2,230,17 (M/min/mg ptn) x 10-1; CS: 4,510,31 (M/min/mg ptn) x 10-1, p<0,001; CS+P: 2,640,19 (M/min/mg ptn) x 10-1, p<0,05). O inverso ocorreu com a relação GSH/GSSG (Controle+P: 1,0880,17; CS: 0,7360,07, p<0,05; CS+P: 1,2580,10, p<0,05). Os níveis de MDA (Controle+P: 0,2660,05 nMol/mg ptn; CS: 0,940,076 nMol/mg ptn, p<0,001; CS+P: 0,4980,06 nMol/mg ptn, p<0,01) e de nitrito (Controle+P: 50,014,19 Mol/mg ptn; CS: 108,77,73 Mol/mg ptn, p<0,001; CS+P: 58,843,42 nMol/mg ptn, p<0,01) estavam aumentados no CS que em outros grupos. A expressão de TNF-α foi observada no grupo CS. O tratamento da P apresentou ação anti-inflamatória e antioxidante em macrófagos e em camundongos expostos à fumaça de cigarro, possivelmente pela ação dos polifenóis presentes nela
Resumo:
In this paper, a new blind and readable H.264 compressed domain watermarking scheme is proposed in which the embedding/extracting is performed using the syntactic elements of the compressed bit stream. As a result, it is not necessary to fully decode a compressed video stream both in the embedding and extracting processes. The method also presents an inexpensive spatiotemporal analysis that selects the appropriate submacroblocks for embedding, increasing watermark robustness while reducing its impact on visual quality. Meanwhile, the proposed method prevents bit-rate increase and restricts it within an acceptable limit by selecting appropriate quantized residuals for watermark insertion. Regarding watermarking demands such as imperceptibility, bit-rate control, and appropriate level of security, a priority matrix is defined which can be adjusted based on the application requirements. The resulted flexibility expands the usability of the proposed method.
Resumo:
In this paper, an improved video encryption method for encrypting the sign bit of motion vectors is proposed based on H.264/AVC, which belongs to selective encryption. This method improves upon previous work involving the sign bit encryption of motion vectors by ensuring the four candidates for the encrypted motion vectors are always located in two orthogonal lines. The improved method can provide a much more effective scrambling effect while keeping the encrypted stream format-compliant and the compression ratio unchanged. The combination of the proposed method with encryption of intra prediction modes can further enhance the scrambling effect, especially for the first few frames which are left clear when only the motion vectors are encrypted.