Skin Color Detection Algorithm Research Articles

Hand Gesture Recognition System Based on Indian Sign Language Using SVM and CNN

Hand gestures (HGs) have been extensively acknowledged as a communication standard for the deaf and dumb school system in the Indian sign language (ISL). We suggested an automated Indian sign language recognition (ISLR) method for English in this study. Initially, the hand region is segmented using the Grasshopper optimization algorithm (GOA) based on the skin color model method. The segmentation’s efficacy is further assessed using three techniques: GOA-based skin color detection algorithm (SCDA), particle swarm optimization-based SCDA (PSO-SCDA), and artificial bee colony-based SCDA (ABC-SCDA). Finally, a database for the retrieved motions representing distinct English alphabets is created. For gesture recognition, the system is trained, and a template-based matching approach is used. The classification technique employs the support vector machine (SVM) and the convolution neural network (CNN). Our suggested recognition method achieved the highest accuracy of 97.85% with GOA-SCDA, compared to 89.29% and 93.96% with PSO-SCDA and ABC-SCDA, respectively. Furthermore, CNN outperformed SVM in classification, with a high accuracy of 99.2% and precision of 81.8%.

Facial region detection robust to changing backgrounds

Background/Objectives: These days, many studies have actively been conducted on intelligent robots capable of providing human friendly service. To make natural interaction between humans and robots, it is required to develop the mobile robot-based technology of detecting human facial regions robustly in dynamically changing real backgrounds.Methods/Statistical analysis: This paper proposes a method for detecting facial regions adaptively through the mobile robot-based monitoring of backgrounds in a dynamic real environment. In the proposed method, a camera-object distance and a color change in object background are monitored, and thereby the skin color extraction algorithm most suitable for the measured distance and color is applied. In the face detection step, if the searched range is valid, the most suitable skin color detection method is selected so as to detect facial regions.Findings: To sum up the experimental results, algorithms have a difference in performance depending on a distance and a background color. Overall, the algorithms using neural network showed stable results. The algorithm using Kismet had a good perception rate for the ground truth part of an original image, and a skin color detection rate was greatly influenced by pink and yellow background colors similar to a skin tone, and consequently an incorrect perception rate of background was considerably high. With regard to each algorithm performance depending on a distance, the closer a distance with an object was to 320cm, the more an incorrect perception rate of a background sharply increased. To analyze the performance of each skin color detection algorithm applied to face detection, we examined how much a skin color of an original image was detected by each algorithm. For a skin color detection rate, after the ground truth for the skin of an original image, the number of pixels of the skin color detected by each algorithm was calculated. In this case, the ground truth means a range of the skin color of an original image to detect.Improvements/Applications: We expect that the proposed approach of detecting facial regionsin a dynamic real environment will be used in a variety of application areas related to computer vision and image processing.

Combining Convolutional and Recurrent Neural Networks for Human Skin Detection

Skin detection from images, typically used as a preprocessing step, has a wide range of applications such as dermatology diagnostics, human computer interaction designs, and etc. It is a challenging problem due to many factors such as variation in pigment melanin, uneven illumination, and differences in ethnicity geographics. Besides, age and gender introduce additional difficulties to the detection process. It is hard to determine whether a single pixel is skin or nonskin without considering the context. An efficient traditional hand-engineered skin color detection algorithm requires extensive work by domain experts. Recently, deep learning algorithms, especially convolutional neural networks (CNNs), have achieved great success in pixel-wise labeling tasks. However, CNN-based architectures are not sufficient for modeling the relationship between pixels and their neighbors. In this letter, we integrate recurrent neural networks (RNNs) layers into the fully convolutional neural networks (FCNs), and develop an end-to-end network for human skin detection. In particular, FCN layers capture generic local features, while RNN layers model the semantic contextual dependencies in images. Experimental results on the COMPAQ and ECU skin datasets validate the effectiveness of the proposed approach, where RNN layers enhance the discriminative power of skin detection in complex background situations.

Skin Segmentation Using YUV and RGB Color Spaces

Skin detection is used in many applications, such as face recognition, hand tracking, and human-computer interaction. There are many skin color detection algorithms that are used to extract human skin color regions that are based on the thresholding technique since it is simple and fast for computation. The efficiency of each color space depends on its robustness to the change in lighting and the ability to distinguish skin color pixels in images that have a complex background. For more accurate skin detection, we are proposing a new threshold based on RGB and YUV color spaces. The proposed approach starts by converting the RGB color space to the YUV color model. Then it separates the Y channel, which represents the intensity of the color model from the U and V channels to eliminate the effects of luminance. After that the threshold values are selected based on the testing of the boundary of skin colors with the help of the color histogram. Finally, the threshold was applied to the input image to extract skin parts. The detected skin regions were quantitatively compared to the actual skin parts in the input images to measure the accuracy and to compare the results of our threshold to the results of other's thresholds to prove the efficiency of our approach. The results of the experiment show that the proposed threshold is more robust in terms of dealing with the complex background and light conditions than others.

An Improved Skin Color Detection Algorithm Model

The main objective of designing skin color model is to determine whether the pixel is skin color pixels and generate the skin color mask images. The paper discusses the choise of color space and skin color model designing in skin color detection system, analysis the problems often needed to solve in it and put forward an improved skin color detection algorithm model based on ellipse boundary. The skin color detection experiment is completed. The result of experiment shows the skin color detection algorithm model is good.

배경의 변화에 따른 피부색상 검출 알고리즘의 성능 비교

Accurately extracting skin color regions is very important in various areas such as face recognition and tracking, facial expression recognition, adult image identification, health-care, and so forth. In this paper, we evaluate the performances of several skin color detection algorithms in indoor environments by changing the distance between the camera and the object as well as the background colors of the object. The distance is from 60cm to 120cm and the background colors are white, black, orange, pink, and yellow, respectively. The algorithms that we use for the performance evaluation are Peer algorithm, NNYUV, NNHSV, LutYUV, and Kimset algorithm. The experimental results show that NNHSV, NNYUV and LutYUV algorithm are stable, but the other algorithms are somewhat sensitive to the changes of backgrounds. As a result, we expect that the comparative experimental results of this paper will be used very effectively when developing a new skin color extraction algorithm which are very robust to dynamic real environments.