Region-based image retrieval using region of interest (ROI) according to incremental frame and clustering color image

Content-based image retrieval involves the extraction of global feature images for their retrieval performance in large image databases. Extraction of global features image cause problem of the semantic gap between the high-level meaning and low-level visual features images. In this study RBIR, Region of Interest Based (ROI) Image Retrieval Using Incremental Frame of Color Image was proposed. It combines several methods, including filtering process, image partitioning using clustering and incremental frame formation, complementation law of theory set to generate ROI, NROI, or ER of the region. The concept of weighting as well as a significant query is also incorporated as a query strategy. Extensive experiments were also conducted on the Wang database and the color model selected was the CIE lab. Experimental results show the proposed method is efficient in image retrieval. The performance of the proposed method shows a better average IPR value of 3.51% compared to RGB and 22.92% with the HSV color model. Meanwhile, it also performs better by 36%, 5%, and 24% compared to methods CH (8,2,2), CH (8,3,3), and CH (16,4,4).

Research Based on Visual Sensors and VR in the Field of Visual Culture

Visual sensors provide us with a wealth of multimedia data; one of the core parts of VR technology is to present people with a real and immersive simulation environment; the application of this technology makes the human narrative to achieve visual transformation and further strengthens the central position of visual culture. This paper studies the field of visual culture from two aspects of visual sensor and VR. In this paper, the HSV color model models task processing problems in visual sensor networks using separable load theory and studies ways to seek optimal task scheduling strategies to minimize the completion time of the visual task. This paper mainly uses literature research method, questionnaire survey method, and statistical analysis method to design a model image output process to explore the application of visual culture under the development of visual sensor and VR and analyzes the development status of film, game, and cultural tourism industry VR in the field of visual culture. Questionnaire data show that compared with traditional video games, more people chose games under the application of VR technology for entertainment; in VR game experience, more than 70% are satisfied with game color and style, operation form, scene design, and experience of VR game itself, which shows to some extent the role of visual sensor and VR in visual culture.

Robot Manipulator Control with Inverse Kinematics PD-Pseudoinverse Jacobian and Forward Kinematics Denavit Hartenberg

This paper presents the development of vision-based robotic arm manipulator control by applying Proportional Derivative-Pseudoinverse Jacobian (PD-PIJ) kinematics and Denavit Hartenberg forward kinematics. The task of sorting objects based on color is carried out to observe error propagation in the implementation of manipulator on real system. The objects image captured by the digital camera were processed based on HSV-color model and the centroid coordinate of each object detected were calculated. These coordinates are end effector position target to pick each object and were placed to the right position based on its color. Based on the end effector position target, PD-PIJ inverse kinematics method was used to determine the right angle of each joint of manipulator links. The angles found by PD-PIJ is the input of DH forward kinematics. The process was repeated until the square end effector reached the target. The experiment of model and implementation to actual manipulator were analyzed using Probability Density Function (PDF) and Weibull Probability Distribution. The result shows that the manipulator navigation system had a good performance. The real implementation of color sorting task on manipulator shows the probability of success rate cm is 94.46% for euclidian distance error less than 1.2 cm.

Recognizing Human Emotional State from Audiovisual Signals

In this work, we investigate the recognition of human emotional states from audiovisual signals. We extract prosodic, Mel-frequency Cepstral Coeffieient (MFCC), and formant frequency features to represent the audio characteristic of the emotional speech. A face detection scheme based on HSV color model is used to detect the face from the background. The facial expressions are represented by Gabor wavelet features. We perform feature selection by using the stepwise method based on Mahalanobis distance. The selected features are used to classify the emotional data into their corresponding classes. Different classification algorithms including Gaussian Mixture Model (GMM), K-nearest neighbours(K-NN), Neural Network (NN), and Fisher's Linear Discriminant Analysis (FLDA) are compared in this study. An adaptive multi-classifier scheme involving the analysis of individual class and combinations of different classes is proposed. Our recognition system is tested over a language independent database. The proposed FLDA-based multi-classifier scheme achieves the best overall and individual class recognition accuracy.

Recognizing Human Emotional State from Audiovisual Signals

In this work, we investigate the recognition of human emotional states from audiovisual signals. We extract prosodic, Mel-frequency Cepstral Coeffieient (MFCC), and formant frequency features to represent the audio characteristic of the emotional speech. A face detection scheme based on HSV color model is used to detect the face from the background. The facial expressions are represented by Gabor wavelet features. We perform feature selection by using the stepwise method based on Mahalanobis distance. The selected features are used to classify the emotional data into their corresponding classes. Different classification algorithms including Gaussian Mixture Model (GMM), K-nearest neighbours(K-NN), Neural Network (NN), and Fisher's Linear Discriminant Analysis (FLDA) are compared in this study. An adaptive multi-classifier scheme involving the analysis of individual class and combinations of different classes is proposed. Our recognition system is tested over a language independent database. The proposed FLDA-based multi-classifier scheme achieves the best overall and individual class recognition accuracy.

Flame Detection Based on Deep Learning Using HSV Color Model and Corner Detection Algorithm

Recently, many image classification or object detection models that use deep learning techniques have been studied; however, in an actual performance evaluation, flame detection using these models may achieve low accuracy. Therefore, the flame detection method proposed in this study is image pre-processing with HSV color model conversion and the Harris corner detection algorithm. The application of the Harris corner detection method, which filters the output from the HSV color model, allows the corners to be detected around the flame owing to the rough texture characteristics of the flame image. These characteristics allow for the detection of a region of interest where multiple corners occur, and finally classify the flame status using deep learning-based convolutional neural network models. The flame detection of the proposed model in this study showed an accuracy of 97.5% and a precision of 97%.

LATE BLIGHT SEGMENTATION IN TOMATOES USING SPATIAL DOMAIN METHODS AND COLOUR ANALYSIS

Agricultural productivity is an important factor in a country's economic development. Therefore, the diagnosis of plant diseases is a field of research of great importance for the agricultural sector since it allows us to help recommend strategies to prevent the spread of diseases, thus reducing economic losses. Currently, with the rise of computer systems, computer systems have been developed that allow computer assisted diagnosis in different fields of research, including the agricultural sector. Since late blight is one of the main diseases due to its high incidence and severity, this paper proposes a methodology to obtain late blight segmentation in tomato leaf images through image analysis and color analysis using the HSV color model. The proposed methodology was applied to a set of publicly available PlantVillage images, to which late blight segmentation was obtained.