Detecting cartoons: a case study in automatic video-genre classification

Ianeva, T.; Vries, Arjen P. de; Röhrig, Hein

doi:10.1109/icme.2003.1220951

Cited by 41 publications

(32 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the prior works, several classification systems based on different types of image features have been developed. In [1] the classification problem was addressed by modeling the characteristics of cartoon. The features are extracted from the color saturation, color histogram, edge histogram, compression ratio, pattern spectrum and the ratio of image pixels with brightness greater than a threshold.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Identifying Computer Graphics using HSV Color Model and Statistical Moments of Characteristic Functions

Chen

Shi

Xuan

2007

Multimedia and Expo, 2007 IEEE International Conference On

141

View full text Add to dashboard Cite

Computer graphics generated by advanced rendering software come to appear so photorealistic that it has become difficult for people to visually differentiate them from photographic images. Consequently, modern computer graphics may be used as a convincing form of image forgery. Therefore, identifying computer graphics has become an important issue in image forgery detection. In this paper, a novel approach to distinguishing computer graphics from photographic images is introduced. The statistical moments of characteristic function of the image and wavelet subbands are used as the distinguishing features. In addition, we investigate the influence of different image color representations on the feature effectiveness. Specifically, the efficiency of using RGB and HSV color models is investigated. The experiments have shown that the features extracted from HSV color space, which decouples brightness from chromatic components, have demonstrated better performance than that from RGB color model.

show abstract

Section: Introductionmentioning

confidence: 99%

“…In [3], the joint spatial-color patch statistics, fractal geometry and differential geometry are formed in RGB space. In [1], some features are collected in HSV (Hue, Saturation, Value (brightness)) color space. In this paper, the proposed method constructs all features in the HSV color space.…”

Section: Introductionmentioning

confidence: 99%

Identifying Computer Graphics using HSV Color Model and Statistical Moments of Characteristic Functions

Chen

Shi

Xuan

2007

Multimedia and Expo, 2007 IEEE International Conference On

141

View full text Add to dashboard Cite

show abstract

“…Simple visual descriptors have been proposed to identify nonphotorealistic computer graphics [1,36,72,26,6]. For example, Ianeva et al [26] observed that the cartoonist graphics has characteristics of saturated and uniform colors, strong and distinct lines, and limited number of colors. They devised a computer graphics detection method that used image features such as the average color saturation, the ratio of image pixels with brightness greater than a threshold, the Hue-Saturation-Value (HSV) color histogram, the edge orientation and strength histogram, the compression ratio, and the distribution of image region sizes.…”

Section: Methods Using Visual Descriptorsmentioning

confidence: 99%

Discrimination of Computer Synthesized or Recaptured Images from Real Images

Chang

2012

Digital Image Forensics

View full text Add to dashboard Cite

An image that appears to be a photograph may not necessarily be a normal photograph as we know it. For example, a photograph-like image can be rendered by computer graphics instead of being taken by a camera or it can be a photograph of an image instead of a direct photograph of a natural scene. What is really different between these photographic appearances is their underlying synthesis processes. Not being able to distinguish these images poses real social risks, as it becomes harder to refute claims of child pornography as non-photograph in the court of law and easier for attackers to mount an image or video replay attack on biometric security systems. This motivates digital image forensics research on distinguishing these photograph-like images from true photographs. In this chapter, we present the challenges, technical approaches, system design, and other practical issues in tackling this multimedia forensics problem. We will also share a list of the open resources and the potential future research directions in this area of research which we hope readers will find useful. MotivationsSince the ancient times of the Greeks and Romans, artists have been playing with special painting techniques for inducing visual illusion where objects in a painting appear to be real and immersed in the real surrounding. Trompe l'oeil, the name for such visual artistry, literally means deceiving the eye. For example, the painting entitled Escaping Criticism created by Pere Borrell del Caso in 1874 depicts a person climbing out of the painting with a make-believe quality ( Figure 1a); the mural painting on the facade of the Saint-Georges Theater created by a painter Dominique Trompe l'oeil makes believe with not only the photorealistic quality in the painting, it also exploits the visual gullibility of human observers through immersion into the real surrounding. Such adversarial nature of trompe l'oeil aptly mirrors that of digital image forensics where the intention to deceive is present. While the deceptive intent is from humans, the photorealism of an image which takes many years of practice for conventional artists to master can be easily produced by laymen with modern technology. Physics-based computer graphics is capable of rendering photorealistic images that emulate images of real three-dimensional scenes, a great advancement from the two-dimensional graphics like cartoon which was popular in the early days of computer graphics. It has been shown that a scene with diffuse reflectance can be realistically rendered to the extent that the rendered scene radiance is close to that of a real scene and perceptually indistinguishable for humans [43].A photograph is photorealistic by definition. Recapturing a photograph with a camera when displayed in good quality on a paper or screen preserves the photorealism of the photograph if perspective distortion is minimized 1 . We refer to such types of images as recaptured or rephotographed image. The modern artist Richard Prince pioneered a unique art form of rephotographing...

show abstract

“…We compare the 192D geometry feature against the 216D wavelet feature [7] and the 108D feature obtained from modelling the characteristics of the general (i.e., including both photorealistic and non-photorealistic) computer graphics [5] (henceforth the cartoon feature). For a fair comparison, we compute the wavelet feature on the entire image (for a better performance), rather than just on the central 256×256-pixel region of an image, as described in [7].…”

Section: Experiments and Discussionmentioning

confidence: 99%

“…Despite the fact that classification of photographic images and computer graphics has been applied for improving the image and video retrieval performance [5,13], classification of photographic images (PIM) and photorealistic computer graphics (PRCG) is a new problem. The work in [7] takes advantage of the wavelet-based natural image statistics, and extract the first four order statistics of the in-subband coefficients and those of the cross-subband coefficient prediction errors as features for classifying PIM and PRCG.…”

Section: Introductionmentioning

confidence: 99%

Physics-motivated features for distinguishing photographic images and computer graphics

Chang

Hsu

et al. 2005

Proceedings of the 13th Annual ACM International Conference on Multimedia

150

118

View full text Add to dashboard Cite

The increasing photorealism for computer graphics has made computer graphics a convincing form of image forgery. Therefore, classifying photographic images and photorealistic computer graphics has become an important problem for image forgery detection. In this paper, we propose a new geometrybased image model, motivated by the physical image generation process, to tackle the above-mentioned problem. The proposed model reveals certain physical differences between the two image categories, such as the gamma correction in photographic images and the sharp structures in computer graphics. For the problem of image forgery detection, we propose two levels of image authenticity definition, i.e., imaging-process authenticity and scene authenticity, and analyze our technique against these definitions. Such definition is important for making the concept of image authenticity computable. Apart from offering physical insights, our technique with a classification accuracy of 83.5% outperforms those in the prior work, i.e., wavelet features at 80.3% and cartoon features at 71.0%. We also consider a recapturing attack scenario and propose a counter-attack measure. In addition, we constructed a publicly available benchmark dataset with images of diverse content and computer graphics of high photorealism.

show abstract

Detecting cartoons: a case study in automatic video-genre classification

Cited by 41 publications

References 12 publications

Identifying Computer Graphics using HSV Color Model and Statistical Moments of Characteristic Functions

Identifying Computer Graphics using HSV Color Model and Statistical Moments of Characteristic Functions

Discrimination of Computer Synthesized or Recaptured Images from Real Images

Physics-motivated features for distinguishing photographic images and computer graphics

Contact Info

Product

Resources

About