No abstract

Abstract. In recent years, there has been a dramatic increase in the number of images captured by users. This is due to the wide availability of digital cameras and mobile phones which are able to capture and transmit images. Simultaneously, image-editing applications have become more usable, and a casual user can easily improve the quality of an image or change its content. The most common type of image modication is cloning, or copy-move forgery (CMF), which is easy to implement and dicult to detect. In most cases, it is hard to detect CMF with the naked eye and many possible manipulations (attacks) can be used to make the doctored image more realistic. In CMF, the forger copies part(s) of the image and pastes them back into the same image. One possible transformation is rotation, where an object is copied, rotated and pasted. Rotation-invariant features need to be used to detect Copy-Rotate-Move (CRM) forgery. In this paper we presented three contributions. First, a new technique to detect CMF is developed, using Dense Scale-Invariant Feature Transform (DSIFT). Second, a new improved DSIFT descriptor is implemented which is more robust to rotation than Zernike moments. Third, a new method to remove false matching is proposed. Extensive experiments have been conducted to train, evaluate and test the algorithms, the new feature vector and the suggested method to remove false matching. We show that the proposed method can detect forgery in images with blurring, brightness change, colour reduction, JPEG compression, variations in contrast and added noise.

Road signs are so important because they help preserve safe driving conditions; they also influence the safety of drivers and pedestrians. Without these signs, no one would know the driving speed limit, on which direction to drive down a road, any upcoming hazard, or whether they are approaching a merge. It would be chaotic to drive in such situations. Moreover, these signs help new drivers to find their way in the absence of navigators. Therefore, traffic sign recognition takes a critical place in computer vision applications to develop an effective algorithm. In order to tackle this challenge, we proposed the use of Multi-language Traffic Sign Detection and Classification. One of our contributions in this work is that, instead of using the standard grayscale image, we used the RGB colored image. This image is converted into the 2D highest-level grayscale image using the largest values of each pixel in the RGB channels. The novel generated image has the strongest features of the RGB image that make the features distinct and more informative in the classification step. Consider that, in general, the traffic sign has two colors only, the foreground (text location) and background (non-text location). The Maximally Stable Extremal Regions (MSER) used to extract features from the 2D image where the locations of interest are well-identified exclusively by an extremal property of the intensity function in the location and on its outer boundary. The geometrical properties and thinning operations were used to remove the non-text locations. A multi-language OCR was used to understand multi-language. This proposed method has been tested using 240 images which were collected from the Internet and two datasets. The experimental results demonstrated the performance of the proposed method where the traffic sign detected in 92% of the tested images with a very high percentage of localization.

No abstract