Papers by Dipak kumar Ghosh
Nanotechnology Perceptions, 2024
Plant diseases greatly affect agricultural productivity and quality. Effective illness management... more Plant diseases greatly affect agricultural productivity and quality. Effective illness management involves early detection and accurate diagnosis. However, hand identification is laborious and error-prone, which can cause losses. The biggest challenge is developing a reliable plant disease identification system. Plant pictures are complex, making it hard to identify disease symptoms from healthy structures. A high-level framework using powerful image processing to automate disease identification is needed. A reliable method to help farmers and other stakeholders detect and diagnose plant diseases are the goal. This research proposes a high-level deep learning system for plant disease detection, focusing on Kaggle repository classification of different plant illnesses. Advanced techniques including gradient-based Radial Basis Function (RBF) for segmentation and Deep Belief Network (DBN) for feature extraction were used to extract relevant features from plant photos using deep learning models. The classification phase used ResNet-50, known for its ability to understand complicated patterns and identify images reliably. Plants are photographed from several angles in real time under stable lighting. The ResNet-50 CNN method, known for extracting hierarchical features from images, classifies diseases. Independent test photos from the New Plant Diseases Dataset are used to validate algorithm performance. Plant disease identification performance improved by 6.7% to 30% over conventional approaches.
Affine Differential Local Mean ZigZag Pattern for Texture Classification
TENCON 2018 - 2018 IEEE Region 10 Conference
The texture classification is a significant problem in the area of pattern recognition. This work... more The texture classification is a significant problem in the area of pattern recognition. This work proposes a novel Affine Differential Local Mean ZigZag Pattern (ADLMZP) descriptor for texture classification. The proposed method has two manifolds: first Local Mean ZigZag Pattern (LMZP) map is calculated by thresholding the 3 × 3 patch neighbor intensity values with respect to path mean but in a ZigZag weighting fashion, which provides a well discriminated descriptor compared to other local binary descriptors. The local micropattern is obtained by comparing neighbor intensity values with respect to path mean which makes the descriptor robust against noise and illumination variations. Secondly, in order to make it invariant to affine changes, we incorporated an affine differential transformation along with affine gradient magnitude information of a texture image which is differed from Euclidean Gradient. The final ADLMZP descriptor is generated by concatenating the histograms of all Affine Differential Local Mean ZigZag maps. The results are computed over well known KTH-TIPS, Brodatz, and CUReT texture datasets and compared with the state-of-the-art texture classification methods.
arXiv (Cornell University), Jan 9, 2018
In this paper we propose a novel texture descriptor called Fractal Weighted Local Binary Pattern ... more In this paper we propose a novel texture descriptor called Fractal Weighted Local Binary Pattern (FWLBP). The fractal dimension (FD) measure is relatively invariant to scale-changes, and presents a good correlation with human viewpoint of surface roughness. We have utilized this property to construct a scale-invariant descriptor. Here, the input image is sampled using an augmented form of the local binary pattern (LBP) over three different radii, and then used an indexing operation to assign FD weights to the collected samples. The final histogram of the descriptor has its features calculated using LBP, and its weights computed from the FD image. The proposed descriptor is scale invariant, and is also robust in rotation or reflection, and partially tolerant to noise and illumination changes. In addition, the local fractal dimension is relatively insensitive to the bi-Lipschitz transformations, whereas its extension is adequate to precisely discriminate the fundamental of texture primitives. Experiment results carried out on standard texture databases show that the proposed descriptor achieved better classification rates compared to the state-of-the-art descriptors.
A Complete Dual-Cross Pattern for Unconstrained Texture Classification
2017 4th IAPR Asian Conference on Pattern Recognition (ACPR)
In order to perform unconstrained texture classification, this paper presents a novel and computa... more In order to perform unconstrained texture classification, this paper presents a novel and computationally efficient texture descriptor called Complete Dual-Cross Pattern (CDCP), which is robust to gray-scale changes and surface rotation. To extract CDCP, at first a gray scale normalization scheme is used to reduce the illumination effect and, then CDCP feature is computed from holistic and component levels. A local region of the texture image is represented by it's center pixel and difference of sign-magnitude transform (DSMT) at multiple levels. Using a global threshold, the gray value of center pixel is converted into a binary code named DCP center (DCP_C). DSMT decomposes into two complementary components: the sign and the magnitude. They are encoded respectively into DCP-sign (DCP_S) and DCP-magnitude (DCP_M), based on their corresponding threshold values. Finally, CDCP is formed by fusing DCP_S, DCP_M and DCP_C features through joint distribution. The invariance characteristics of CDCP are attained due to computation of pattern at multiple levels, which makes CDCP highly discriminative and achieves state-of-the-art performance for rotation invariant texture classification.
In today’s technical world, the intellectual computing of a efficient human-computer interaction ... more In today’s technical world, the intellectual computing of a efficient human-computer interaction (HCI) or human alternative and augmentative communication (HAAC) is essential in our lives. Hand gesture recognition is one of the most important techniques that can be used to build up a gesture based interface system for HCI or HAAC application. Therefore, suitable development of gesture recognition method is necessary to design advance hand gesture recognition system for successful applications like robotics, assistive systems, sign language communication, virtual reality etc. However, the variation of illumination, rotation, position and size of gesture images, efficient feature representation, and classification are the main challenges towards the development of a real time gesture recognition system. The aim of this work is to develop a framework for vision based static hand gesture recognition which overcomes the challenges of illumination, rotation, size and position variation of...
A static hand gesture recognition algorithm using k-mean based radial basis function neural network
2011 8th International Conference on Information, Communications & Signal Processing, 2011
Applied Soft Computing, 2019
This is a PDF file of an article that has undergone enhancements after acceptance, such as the ad... more This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Multimedia Tools and Applications, 2018
Methods based on locally encoded image features have recently become popular for texture classifi... more Methods based on locally encoded image features have recently become popular for texture classification tasks, particularly in the existence of large intra-class variation due to changes in illumination, scale, and viewpoint. Inspired by the theories of image structure analysis, this work proposes an efficient, simple, yet robust descriptor namely local jet pattern (LJP) for texture classification. In this approach, a jet space representation of a texture image is computed from a set of derivatives of Gaussian (DtGs) filter responses up to second order, so-called local jet vectors (LJV), which also satisfy the Scale Space properties. The LJP is obtained by using the relation of center pixel with its' local neighborhoods in jet space. Finally, the feature vector of a texture image is formed by concatenating the histogram of LJP for all elements of LJV. All DtGs responses up to second order together preserves the intrinsic local image structure, and achieves invariance to scale, rotation, and reflection. This allows us to design a discriminative and robust framework for texture classification. Extensive experiments on five standard texture image databases, employing nearest subspace classifier (NSC), the proposed descriptor achieves 100%, 99.92%, 99.75%, 99.16%, and 99.65% accuracy for Outex TC10, Outex TC12, KTH-TIPS, Brodatz, CUReT, respectively, which are better compared to state-of-the-art methods.
Digital Signal Processing, 2018
This work sheds light on real world texture classification problem where textural information pre... more This work sheds light on real world texture classification problem where textural information present at several scales as images are captured from different distances and zooming conditions. Inspired by the concept of mathematical morphology in image analysis, we developed a novel statistical approach to texture representation, which yields more discriminative, simple, efficient, yet robust descriptor called local morphology pattern (Lmp). The proposed approach consists of two parts: first it generates a morphological scale space representation of the texture image, called local shape vector (Lsv) using two commonly used powerful morphology operations namely, "opening" and "closing" at different scale. It perfectly localizes and well preserves the contours information of the objects. Then Lmp is formed by utilizing the relation of center pixel with its neighboring pixels in the scale space. The encoding of Lmp is done via uniform pattern scheme. The normalized histogram of Lmp (Hlmp) is computed by concatenating the histograms of Lmp at different scales. It is observed that Lmp is invariant under local Lipschitz transform and its extension is adequate to precisely differentiate between the fundamental texture primitives. Lipschitz is a very conventional transform which incorporates rotation, translation, projective transformation and also changes in viewpoint. Extensive experiments of texture classification on the benchmark texture databases Outex_TC-00010 (Outex_TC10), Outex_TC-00012 (Outex_TC12), KTH-TIPS, and UIUC validate that the proposed Lmp descriptor has better or comparable performance with well-known and state-of-the-art methods.
Pattern Recognition Letters, 2018
Local feature descriptors play a key role in texture classification tasks. However, such traditio... more Local feature descriptors play a key role in texture classification tasks. However, such traditional descriptors are deficient to capture the edges and orientations information and local intrinsic structure of images. This letter introduces a simple, new, yet powerful rotation invariant texture descriptor named Local Directional ZigZag Pattern (Ldzp) by ZigZag scanning for effective representation of texture. Here at first we compute the directional edge information, so called local directional edge map (Ldem) of a texture image using the Kirsch compass mask in six different directions. Then Local ZigZag Pattern (Lzp) is extracted from all Ldem images. Basically, the Lzp characterizes the spatial ZigZag structure based on the relation between reference pixel and its adjacent neighboring pixels and is insensitive to the illumination changes. Finally, the uniform pattern histograms are computed from all directional Lzp maps which are concatenated to form the final Ldzp descriptor. Extensive experiments on texture classification shows the proposed Ldzp descriptor achieves state-of-the-art performance in terms of average classification accuracy when applied to the large and well-known benchmark Outex database. We have also shown that Ldzp descriptor is equally powerful for human face recognition.
2015 Fifth International Conference on Communication Systems and Network Technologies, 2015
A hand gesture recognition system has a wide area of application in human computer interaction (H... more A hand gesture recognition system has a wide area of application in human computer interaction (HCI) and sign language. This work proposes a vision-based system for recognition of static hand gesture. It deals with images of bare hands, and allows to recognize gesture in illumination, rotation, position and size variation of gesture images. The proposed system consists of three phases: preprocessing, feature extraction and classification. The preprocessing phase involves image enhancement, segmentation, rotation and filtering process. To obtain a rotation invariant gesture image, a novel technique is proposed in this paper by coinciding the 1 st principal component of the segmented hand gestures with vertical axes. In feature extraction phase, this work extracts localized contour sequences (LCS) and block based features and proposes a novel mixture of features (or combined features) for better representation of static hand gesture. The combined features are applied as input to multiclass support vector machine (SVM) classifier to recognize static hand gesture. The proposed system is implemented and tested on three different hand alphabet databases. The experimental results show that the proposed system able to recognize static gesture with a recognition sensitivity of 99.50%, 93.58% and 98.33% for database I, database II and database III respectively which are better compared to earlier reported methods.
The aim of this paper is to automatically segment the hand gesture from a given image under diffe... more The aim of this paper is to automatically segment the hand gesture from a given image under different luminance conditions and complex backgrounds. The luminance value affects the color component of an image which leads to increase the noise level in the segmented image. This paper proposes a combined model of two color spaces i.e., HSI, YCbCr and morphological operations with labeling to improve the segmentation performance of color hand gesture from complex backgrounds in terms of completeness and correctness. The proposed color model separates the chrominance and luminance components of the image. The performance of the proposed method is demonstrated through simulation and the experimental results reveal that proposed method provides better performance accuracy compared to the HSI and YCbCr methods individually in terms of correctness and completeness.
2013 15th International Conference on Advanced Computing Technologies (ICACT), 2013
Mammographic screening is the most effective procedure for the early detection of breast cancers.... more Mammographic screening is the most effective procedure for the early detection of breast cancers. However, typical diagnostic signs such as masses are difficult to detect as mammograms are low-contrast noisy images. This paper proposes a systematic method for the detection of suspicious lesions in digital mammograms based on undecimated wavelet transform and adaptive thresholding techniques. Undecimated wavelet transform is used here to generate a multiresolution representation of the original mammogram. Adaptive global and local thresholding techniques are then applied to segment possible malignancies. The segmented regions are enhanced by using morphological filtering and seeded region growing. The proposed method is evaluated on 120 images of the Mammographic Image Analysis Society (MIAS) Mini Mammographic database, that include 89 images having in total 92 lesions. The experimental results show that the proposed method successfully detects 87 of the 92 lesions, performing with a sensitivity of 94.56% at 0.8 false positives per image (FPI), which is better than earlier reported techniques. This shows the effectiveness of the proposed system in detecting breast cancer in early stages.
Advances in Intelligent Systems and Computing, 2014
TexFusionNet: An Ensemble of Deep CNN Feature for Texture Classification
Proceedings of 3rd International Conference on Computer Vision and Image Processing
The texture classification from images is one of the important problems in pattern recognition. S... more The texture classification from images is one of the important problems in pattern recognition. Several hand-designed methods have been proposed in last few decades for this problem. Nowadays, it is observed that the convolutional neural networks (CNN) perform extremely well for the classification task mainly over object and scene images. This improved performance of CNN is caused by the availability of huge amount of images in object and scene databases such as ImageNet. Still, the focus of CNN in texture classification is limited due to non-availability of large texture image data sets. Thus, the trained CNN over Imagenet database is used for texture classification by fine-tuning the last few layers of the network. In this paper, a fused CNN (TexFusionNet) is proposed for texture classification by fusing the last representation layer of widely adapted AlexNet and VGG16. On the top of the fused layer, a fully connected layer is used to generate the class score. The categorical cross-entropy loss is used to generate the error during training, which is used to train the added layer after the fusion layer. The results are computed over several well-known Brodatz, CUReT, and KTH-TIPS texture data sets and compared with the state-of-the-art texture classification methods. The experimental results confirm outstanding performance of the proposed TexFusionNet architecture for texture classification.
This study demonstrates the development of vision based static hand gesture recognition system us... more This study demonstrates the development of vision based static hand gesture recognition system using web camera in real-time applications. The vision based static hand gesture recognition system is developed using the following steps: preprocessing, feature extraction and classification. The preprocessing stage consists of illumination compensation, segmentation, filtering, hand region detection and image resize. This study proposes a discrete wavelet transform (DWT) and Fisher ratio (F-ratio) based feature extraction technique to classify the hand gestures in an uncontrolled environment. This method is not only robust towards distortion and gesture vocabulary, but also invariant to translation and rotation of hand gestures. A linear support vector machine is used as a classifier to recognise the hand gestures. The performance of the proposed method is evaluated on two standard public datasets and one indigenously developed complex background dataset for recognition of hand gestures. All above three datasets are developed based on American Sign Language (ASL) hand alphabets. The experimental result is evaluated in terms of mean accuracy. Two possible real-time applications are conducted, one is for interpretation of ASL sign alphabets and another is for image browsing.
COVID-19 pandemic has had a dramatic impact on our daily lives, disrupting global trade and trans... more COVID-19 pandemic has had a dramatic impact on our daily lives, disrupting global trade and transportation. Protecting one's face with a mask has become the new normal. In the present new normal situation and in near future also, customers must wear masks properly in many public services. Therefore, face mask identification has become a necessary and challenging stuff in the present new normal scenario. In this paper, a computer vision-based method using proposed convolution neural network (CNN), named as FusionNet is introduced to correctly detect the face mask. The proposed FusionNet architecture is constructed by combining of VGG16Net with MobileNetV2. This approach accurately recognizes the face mask from a video sequence and determines whether a person wearing a mask or not. The method is evaluated on the dataset which consist of total 400 images (200 images with mask and 200 images without mask) captured from different social gatherings. The proposed method can also be used to detect multiple faces with or without mask in a single video frame and achieves 99.75% accuracy which is better compared to individual performance of using standard VGG16Net and MobileNetV2.
The accurate classification of static hand gestures is a vital role to develop a hand gesture rec... more The accurate classification of static hand gestures is a vital role to develop a hand gesture recognition system which is used for human-computer interaction (HCI) and for human alternative and augmentative communication (HAAC) application. A vision-based static hand gesture recognition algorithm consists of three stages: preprocessing, feature extraction and classification. The preprocessing stage involves following three sub-stages: segmentation which segments hand region from its background images using a histogram based thresholding algorithm and transforms into binary silhouette; rotation that rotates segmented gesture to make the algorithm, rotation invariant; filtering that effectively removes background noise and object noise from binary image by morphological filtering technique. To obtain a rotation invariant gesture image, a novel technique is proposed in this paper by coinciding the 1 st principal component of the segmented hand gestures with vertical axes. A localized contour sequence (LCS) based feature is used here to classify the hand gestures. A k-mean based radial basis function neural network (RBFNN) is also proposed here for classification of hand gestures from LCS based feature set. The experiment is conducted on 500 train images and 500 test images of 25 class grayscale static hand gesture image dataset of Danish/international sign language hand alphabet. The proposed method performs with 99.6% classification accuracy which is better than earlier reported technique.
This paper proposes a simple and effective texture recognition method that uses a new class of je... more This paper proposes a simple and effective texture recognition method that uses a new class of jet texton learning. In this approach, first a Jet space representation of the image is derived from a set of derivative of Gaussian (DtGs) filter responses upto 2nd order (R 6), so called local jet vector (Ljv), which satisfies the scale space properties, where the combinations of local jets preserve the intrinsic local structure of the image in a hierarchical way and are invariant to image translation, rotation and scaling. Next, the jet textons dictionary is learned using K-means clustering algorithm from DtGs responses, followed by a contrast Weber law normalization pre-processing step. Finally, the feature distribution of jet texton is considered as a model which is utilized to classify texture using a nonparametric nearest regularized subspace (Nrs) classifier. Extensive experiments on three large and well-known benchmark database for texture classification like KTH-TIPS, Brodatz and CUReT show that the proposed method achieves state-of-the-art performance, especially when the number of available training samples is limited. The source code of complete system is made publicly available at https://github.com/swalpa/JetTexton.
A vision-based static hand gesture recognition method which consists of preprocessing, feature ex... more A vision-based static hand gesture recognition method which consists of preprocessing, feature extraction, feature selection and classification stages is presented in this work. The preprocessing stage involves image enhancement, segmentation, rotation and filtering. This work proposes an image rotation technique that makes segmented image rotation invariant and explores a combined feature set, using localized contour sequences and block-based features for better representation of static hand gesture. Genetic algorithm is used here to select optimized feature subset from the combined feature set. This work also proposes an improved version of radial basis function (RBF) neural network to classify hand gesture images using selected combined features. In the proposed RBF neural network, the centers are automatically selected using k-means algorithm and estimated weight matrix is recursively updated, utilizing least-mean-square algorithm for better recognition of hand gesture images. The comparative performances are tested on two indigenously developed databases of 24 American sign language hand alphabet. Keywords American sign language (ASL) hand alphabet • Combined feature • Genetic algorithm (GA) • Hand gesture recognition • Least-mean-square (LMS) algorithm • Radial basis function (RBF) neural network B Dipak Kumar Ghosh
Uploads
Papers by Dipak kumar Ghosh