To search, Click below search items.


All Published Papers Search Service


DNN Based Ant Colony Optimization for Video Tampering Detection


Siva Prasad Patnayakuni


Vol. 22  No. 9  pp. 832-839


Lately, videotape tampering process becomes easier due to the rapid-fire advancements in stoner-friendly editing software and multimedia technology(e.g., Mokey by Imagineer Systems, and Photoshop and Premiere by Adobe). This technologies may largely tamper the original images, so that the followership gets mislead. First, an image is divided into blocks of different sizes by a rate-distortion-based modified horizontal-vertical partition scheme. Statistical redundancy of quantized DCT portions of each image block is reduced by a bit-plane dynamical arithmetical coding with a sophisticated environment modeling. The compressed frames are also segmented by introducing Watershed segmentation fashion. Then, the region- grounded approach where the target structure is regarded as a homogeneous region which is determined by a hunt process guided by applicable criteria for unity. This unity features are also handed to DNN( deep neural network) for phony discovery. In this videotape forensic process, DNN classifier is included for phony discovery. The CNN classifier is included in colorful being phony discovery ways. But, in our work we include DNN because it contains number of retired layers which give accurate results for this phony discovery process. To ameliorate the DNN performance, Ant Colony Optimization (ACO) algorithm is introduced in this phony discovery fashion. Every niche and corner of this world we can suitable to find the surveillance cameras for security purpose. But, some fraudsters perform phonies in this recorded videos for their own benefits. To identify this, a lot of phony discovery ways are coming into actuality. So in this work, we introduce the DNN grounded ACO to perform the phony discovery in images. This perpetration is reused in python simulation platform. The parametric evaluations are taken in terms of F1- Score, accuracy, Precision, Recall and the Experimental results will give bettered performance in videotape phony discovery.


DCT Compression, Least Mean square, watershed segmentation, Deep neural network, optimization.