Shadow Detection and Removal Based on YCbCr Color Space

Kaushik Deb,Ashraful Huq Suny 한국산학기술학회 2014 SmartCR Vol.4 No.1

Shadows in an image can reveal information about the object’s shape and orientation, and even about the light source. Thus shadow detection and removal is a very crucial and inevitable task of some computer vision algorithms for applications such as image segmentation and object detection and tracking. This paper proposes a simple framework using the luminance, chroma: blue, chroma: red (YCbCr) color space to detect and remove shadows from images. Initially, an approach based on statistics of intensity in the YCbCr color space is proposed for detecting shadows. After the shadows are identified, a shadow density model is applied. According to the shadow density model, the image is segmented into several regions that have the same density. Finally, the shadows are removed by relighting each pixel in the YCbCr color space and correcting the color of the shadowed regions in the red-green-blue (RGB) color space. The most salient feature of our proposed framework is that after removing shadows, there is no harsh transition between the shadowed parts and non-shadowed parts, and all the details in the shadowed regions remain intact. Various shadow images were used with a variety of conditions (i.e. outdoor and semi-indoor) to test the proposed framework, and results are presented to prove its effectiveness.

DCT and DWT Based Robust Audio Watermarking Scheme for Copyright Protection

Deb, Kaushik,Rahman, Md. Ashikur,Sultana, Kazi Zakia,Sarker, Md. Iqbal Hasan,Chong, Ui-Pil The Korea Institute of Convergence Signal Processi 2014 융합신호처리학회 논문지 (JISPS) Vol.15 No.1

Digital watermarking techniques are attracting attention as a proper solution to protect copyright for multimedia data. This paper proposes a new audio watermarking method based on Discrete Cosine Transformation (DCT) and Discrete Wavelet Transformation (DWT) for copyright protection. In our proposed watermarking method, the original audio is transformed into DCT domain and divided into two parts. Synchronization code is applied on the signal in first part and 2 levels DWT domain is applied on the signal in second part. The absolute value of DWT coefficient is divided into arbitrary number of segments and calculates the energy of each segment and middle peak. Watermarks are then embedded into each middle peak. Watermarks are extracted by performing the inverse operation of watermark embedding process. Experimental results show that the hidden watermark data is robust to re-sampling, low-pass filtering, re-quantization, MP3 compression, cropping, echo addition, delay, and pitch shifting, amplitude change. Performance analysis of the proposed scheme shows low error probability rates.

Cast Shadow Detection and Removal of Moving Objects from Video Based on HSV Color Space

Kaushik Kaushik,Deb, Animesh Kar,Ashraful Huq Suny 한국산학기술학회 2015 SmartCR Vol.5 No.1

In the process of segmentation and tracking of moving object, shadow area leads to false detection of object. Shadows are also reason for loss of background texture and false connectivity of independent blobs. Hence, we propose a simple method to detect a moving object’s cast shadow and remove the shadow region from video frames. Initially, we store all the background information in reference frame. The next incoming frames with object are compared with this frame. In order to extract the moving object, we used subtraction algorithm. We used homogeneity property by image division, color variation in RGB color space and statistics of intensity in V channel of HSV color space to detect the shadow region. Finally shadow removal is done based on the information from the reference frame. The most noticeable feature of our proposed method is that it can detect shadows both in indoor and outdoor scenarios and there is no harsh transition after removal of the shadow. Color information for both background subtraction and shadow detection to improve object segmentation is ensured in this paper. Experimental results show that the proposed method is simple to understand, can detect and remove shadow and extract the moving object properly.

A VEHICLE LICENSE PLATE DETECTION METHOD FOR INTELLIGENT TRANSPORTATION SYSTEM APPLICATIONS

Deb, Kaushik,Jo, Kang-Hyun Taylor Francis 2009 Cybernetics and systems Vol.40 No.8

<P> Detecting license plates is crucial and inevitable in the vehicle license plate recognition system. In this article, a Hue-Saturation-Intensity (HSI) color model is adopted to select automatically statistical threshold value for detecting candidate regions. The focus of this article is on the implementation of a new method to detect candidate regions when vehicle bodies and license plates (LP) have similar color. The proposed method is able to deal with candidate regions under independent orientation and scale of the plate. For the decomposing candidate regions, predetermined LP alphanumeric characters are used by position histogram to verify and detect vehicle LP regions. Various LP images were used with a variety of conditions to test the proposed method and results proved its effectiveness.</P>

DCT and DWT Based Robust Audio Watermarking Scheme for Copyright Protection

Kaushik Deb,Md. Ashikur Rahman,Kazi Zakia Sultana,Iqbal Hasan Sarker,정의필 한국융합신호처리학회 2014 융합신호처리학회 논문지 (JISPS) Vol.15 No.1

Digital watermarking techniques are attracting attention as a proper solution to protect copyright for multimedia data. This paper proposes a new audio watermarking method based on Discrete Cosine Transformation (DCT) and Discrete Wavelet Transformation (DWT) for copyright protection. In our proposed watermarking method, the original audio is transformed into DCT domain and divided into two parts. Synchronization code is applied on the signal in first part and 2 levels DWT domain is applied on the signal in second part. The absolute value of DWT coefficient is divided into arbitrary number of segments and calculates the energy of each segment and middle peak. Watermarks are then embedded into each middle peak. Watermarks are extracted by performing the inverse operation of watermark embedding process. Experimental results show that the hidden watermark data is robust to re-sampling, low-pass filtering, re-quantization, MP3 compression, cropping, echo addition, delay, and pitch shifting, amplitude change. Performance analysis of the proposed scheme shows low error probability rates.

Korean Vehicle License Plate Extraction Based on HSI Color Model and Histogram

Kaushik Deb,Kang-Hyun Jo 한국자동차공학회 2008 한국자동차공학회 춘 추계 학술대회 논문집 Vol.- No.-

A Low Frequency Band Watermarking with Weighted Correction in the Combined Cosine and Wavelet Transform Domain

Deb, Kaushik,Al-Seraj, Md. Sajib,Chong, Ui-Pil The Korea Institute of Convergence Signal Processi 2013 융합신호처리학회 논문지 (JISPS) Vol.14 No.1

A combined DWT and DCT based watermarking technique of low frequency watermarking with weighted correction is proposed. The DWT has excellent spatial localization, frequency spread and multi-resolution characteristics, which are similar to the theoretical models of the human visual system (HVS). The DCT based watermarking techniques offer compression while DWT based watermarking techniques offer scalability. These desirable properties are used in this combined watermarking technique. In the proposed method watermark bits are embedded in the low frequency band of each DCT block of selected DWT sub-band. The weighted correction is also used to improve the imperceptibility. The extracting procedure reverses the embedding operations without the reference of the original image. Compared with the similar approach by DCT based approach and DWT based approach, the experimental results show that the proposed algorithm apparently preserves superiori mage quality and robustness under various attacks such as JPEG compression, cropping, sharping, contrast adjustments and so on.

A Motion Region Detection and Tracking Method

Kaushik Deb,Sayem Imtiaz,Priyam Biswas 한국산학기술학회 2014 SmartCR Vol.4 No.1

Nowadays, video surveillance is indispensable in security-sensitive areas. Hence, a significant amount of work has been done in this field. This paper proposes a hybrid framework for motion region detection and an appearance-based real-time motion tracking system. Initially, a foreground map is extracted through a process of subtraction from a background model, applying a temporal differencing method. Then, shadow elimination and morphological operations are used to remove noise. Finally, models are initiated for each detected motion region by extracting features such as center of mass and a color correlogram, which are then used for tracking purposes. As the similarity in distances within a certain radius is measured, the probability of confusing objects is reduced considerably, and therefore, performance is optimized significantly. The proposed framework also uses a robust technique to label people within a group. This framework has the capability to work in indoor, semi-outdoor, and even outdoor environments that generate a penumbra shadow, and it handles the groups formed due to occlusion effectively. The framework takes good care of false foreground pixels due to penumbra shadow. Hence, the proposed framework will play a pivotal role in providing security in highly confidential areas.

HSI Color based Vehicle License Plate Detection

Kaushik Deb,Kang-Hyun Jo 제어로봇시스템학회 2008 제어로봇시스템학회 국제학술대회 논문집 Vol.2008 No.10

Vehicle license plate recognition (VLPR) is one of the most important topics of using computer vision and pattern recognition in intelligent transportation systems. In order to recognize a license plate (LP) expeditiously, the location of the LP in most cases, must be detected in the initial step. For this reason, detecting the exact and perfect location of a LP from a vehicle image is considered to be the most important and crucial step of a VLPR system, which greatly affects the recognition process and directly influences the accuracy and speed of entire system. In this paper a HSI color based license plate detection method is proposed. In this method, (a) HSI color model is used for detecting candidate regions and (b) vehicle license plate (VLP) regions are verified and detected by using position histogram. In the proposed method, input vehicle images are converted into HSI color images. Then the candidate regions are found by HSI color model on the basis of using hue, saturation and/or intensity. These candidate regions may include LP regions; geometrical properties of LP are then used for classification. Finally, VLP regions containing predetermined LP alphanumeric character are verified and detected by using position histogram. The proposed method is very effective in coping with different conditions such as poor illumination and varied weather comparing with traditional approaches. Experimental results show that the distance from the vehicle varied according to the camera setup.

Baggage Recognition in Occluded Environment using Boosting Technique

( Tahmina Khanam ),( Kaushik Deb ) 한국인터넷정보학회 2017 KSII Transactions on Internet and Information Syst Vol.11 No.11

Automatic Video Surveillance System (AVSS) has become important to computer vision researchers as crime has increased in the twenty-first century. As a new branch of AVSS, baggage detection has a wide area of security applications. Some of them are, detecting baggage in baggage restricted super shop, detecting unclaimed baggage in public space etc. However, in this paper, a detection & classification framework of baggage is proposed. Initially, background subtraction is performed instead of sliding window approach to speed up the system and HSI model is used to deal with different illumination conditions. Then, a model is introduced to overcome shadow effect. Then, occlusion of objects is detected using proposed mirroring algorithm to track individual objects. Extraction of rotational signal descriptor (SP-RSD-HOG) with support plane from Region of Interest (ROI) add rotation invariance nature in HOG. Finally, dynamic human body parameter setting approach enables the system to detect & classify single or multiple pieces of carried baggage even if some portions of human are absent. In baggage detection, a strong classifier is generated by boosting similarity measure based multi layer Support Vector Machine (SVM)s into HOG based SVM. This boosting technique has been used to deal with various texture patterns of baggage. Experimental results have discovered the system satisfactorily accurate and faster comparative to other alternatives.