For shape-basedfeatureextraction, the process will detect loop, lines, and curves that form Java character. For each of Java character, is divided into two segment, upper segment and lower segment. For each segment, the number of detected loop, line and curve is calculated. Those numbers are features that will be used as input to the Java character recognition. For each of Java character there will be seven features from shape- basedfeatureextraction: number of total loop, number of loop in upper segment, number of loop in lower segment, number of line in upper segment, number of line in lower segment,
The process of SVM classifier is shown in Figure 2. Input image is a single image with foreground green and white background that use the format JPEG, BMP. The image is used as many as 56 images consisting of 36 images watermelon plant leaf and 20 weeds leaf with variety conditions. Then perform preprocessing the image by changing the RGB image into a grayscale image, then perform segmentation using edge detection Canny operator. In this proposed classification framework, there are two features. They are shapefeature and texture feature. For shapefeatureextraction is done using morphological features digital (shown in Table 2), which previously had to be informed of the basic features geometry (shown in Table 1). While the extracted texture features using Gray Level Co-occurrence Matrices (GLCM) were determined by contrast, correlation, energy and homogenecity (see Table 3). The data is then divided into training data and testing data. The training data is used to build the model and the testing data is used to test the model that has been built. Input image is used as training data 1-33 and 34-56 images are used as the data 23 testing. The next stage is the process of classification using SVM method.
Featureextraction has an important role in the field of handwritten recognition, especially in reducing the number of data to be processed. Block averaging is one of featureextraction that based on multiresolution of images. 2D Gaussian filter is one of low-pass filter that can be able to blur the image. By performing block averaging on a blurring image, a set of featureextraction values can be resulted. Based on the experiment, it was shown that 2D Gaussian filter 14x14 with standard deviation 10, can blur 64x64 pixels binary image optimally. In this case, it can show basic shape of the image clearly, which is not too detail nor too blur, where quantitatively it give highest recognition rate. Meanwhile, based on the other experiments it was shown that featureextraction by using block averaging can give insignificant difference in terms of recognition rate performance, if it is compared with featureextraction by using wavelet and DCT.
Abstract—Facial emotional expressions recognition (FEER) is important research fields to study how human beings reflect to environments in affective computing. With the rapid development of multimedia technology especially image processing, facial emotional expressions recognition researchers have achieved many useful result. If we want to recognize the human’s emotion via the facial image, we need to extract features of the facial image. Active Shape Model (ASM) is one of the most popular methods for facial featureextraction. The accuracy of ASM depends on several factors, such as brightness, image sharpness, and noise. To get better result, the ASM is combined with Gaussian Pyramid. In this paper we propose a facial emotion expressions recognizing method based on ASM and Radial Basis Function Network (RBFN). Firstly, facial feature should be extracted to get emotional information from the region, but this paper use ASM method by the reconstructed facial shape. Second stage is to classify the facial emotion expressions from the emotional information. Finally get the model which is matched with the facial feature outline after several iterations and use them to recognize the facial emotional expressions by using RBFN. The experimental result from RBFN classifiers show a recognition accuracy of 90.73% for facial emotional expressions using the proposed method.
Hough Transform maps the points in the image into the parameter space (Hough Transform space) based on a function that defines the shape that wants to be detected, and then takes a vote on an array element called the accumulator array. Hough Trans- form is generally used to perform the extraction of lines, circles or ellipses in the im- age, but in its development, Hough Transform can also be used to the extraction of more complex shapes. Hough Transform is used to detect the straight lines that satisfy the Equation 1 and 2:
From the experimental results, there are several points that could be discussed. In Table 2, it can be seen that for category “Dinosaur” in Figure 5(a) has a precision value of 1.0 (100%). This is because the “Dinosaur” has a background image that is relatively homogeneous, consisting only of one color. It affects the segmentation results in the determination of the selected region for local features extraction. The segmentation results can also affect the determination of the shape which is used as a global feature. Category which also produces precision value 1.0 is “Horse” in Figure 5(b). Although the Horse has a background consisting of a variety of colors, but have a different color combination of other colors in the dataset, as well as texture. Another category is ”Bus”, in this category the Bus has similar shape in every image on this category that can be identify well using the featureshape.
Most important information of the image exists in the image edge, which is the most prominent in the image local changes. It reflects the feature differences within theimage local areas and it is indicated as certain discontinuity of the image information . Edge mainly exists between objective and objective, objective and background as well as area and area (including different colors) and it is animportant foundation of the image analysis such as image segmentation, texture features and shape features . Edge is mainly expressed as the discontinuity of image local characteristics and it contains the relatively severe gray level changes in the image, that is, the place withsingular changes in the signal. The gray level change of singular signal becomes increasingly dramatic along the edge. We generally divide the edge into two types: step and roof, which are indicated in Figure 1 and Figure 2 .
To identify or classify objects in the image, first we must extract features from an image and then use this feature in a pattern to obtain a final grade classifier. Featureextraction is used to identify features that can make a better representation of the object. Not only color and texture, shape or morphology can also be used to extract features. Mathematical morphology is a tool to extract image components that are useful to represent and describe shape of region, such as boundaries, skeletons, and the convex hull. Morphological is related to certain operations that are useful for analyzing the shape of the image so that shape of the object can be recognized.
Pertama-tama melakukan digitalisasi citra yang digambar ke biner, yang kedua adalah preprocessing,dan terakhir melakukan ekstraksi shadow feature sekaligus normalisasi nilai fitur.Ekstraksi ciri dilakukan dengan mengukurpanjang bayangan dari tiap-tiap sisi pada masing-masing area. Namun sebelum diekstrak citra melalui tahap preprocessing dimana dalamnyaditentukan dulu ROI (region
Optical character recognition digunakan untuk menerjemahkan karakter pada citra digital menjadi format teks. Penerapannya yang sederhana membuat algoritma template matching banyak digunakan dalam OCR. Namun, penggunaan algoritma template matcing terhadap OCR masih memiliki tingkat akurasi pengenalan yang kecil. Hal ini terjadi karena terbatasnya template yang disimpan dalam basis data sedangkan model template, yaitu font terus bertambah. Kelebihan algoritma featureextraction yang tidak terbatas dalam model template diharapkan bisa meningkatkan akurasi pengenalan terhadap OCR.
The major failure mechanisms in fibre-reinforced composites are fibre breakage, matrix cracking, fibre–matrix interfacial debonding, delamination and fibre pull-out. Figure 5 shows AE signal of matrix cracking, Figure 5(a) is the waveforms of AE signal and the former 3 IMFs by EMD, it is seen that the dominant frequency range of Hilbert marginal spectrum by HHT is 80- 140kHz, the three-dimensional unite time-frequency chart provide instantaneous frequencies in time-scale of an AE signal in Figure 5(c)，obviously ，frequency range of 80-140kHz is a majority in time-scale. In spite of AE signal is decomposed into several IMFs, the FFT(Fast Fourier Transform) analysis is performed on former 3 IMF waveforms(shown in Figure 6), the dominant frequency range of each IMF is still 80-140kHz approximately, it comes to a conclusion that the frequency feature of this AE signal be supposed to correspond to damage mode of matrix cracking.
Edge detection plays an important role in computer vision and image analysis because edge is the basic characteristic of the image. Edge detection can be the main approach to image analysis and recognition because of the useful and identical information that's contained in edge of the sub - image. Edge detection is a fundamental tool in image processing which are in the areas of feature detection and featureextraction. By applying an edge detection algorithm to an image, the amount of data to be processed may be reducing. Edge detection algorithms are filtering out information that may be regarded as less relevant, but it still preserving the important structural properties of an image. If the edge detection step is successful, the subsequent task of interpreting the information contents in the original image may therefore be substantially simplified.
Hazrati, M. K., & Erfanian, A. (2010). An online EEG-based brain – computer interface for controlling hand grasp using an adaptive probabilistic neural network. Medical Engineering & Physics , vol.32 hlm.730-739. Hsu, Wei-Yen. (2010). EEG-based motor imagery classification using neuro-
However, the major drawback of EMG signals is the poor recognition results under conditions of existing noises especially when the frequency characteristic of noise is random. This noise, artifact and interference in recorded EMG signals are electrode noise, electrode and cable motion artifact, alternating current power line interference, and other noise sources such as a broadband noise from electronic instrument . Some of these noises can be easily removed using filtering methods. Despite that, the interferences of random noise that fall in EMG dominant frequency energy are difficult to be removed. Therefore, by featureextraction, the relevant structures in the EMG signals are highlighted and the noise and unimportant EMG signals are rejected.
Sistem deteksi kualitas tembakau yang dilengkapi kamera yang ditempatkan pada sebuah ruang pengujian dengan warna background hitam telah mampu melakukan proses klasifikasi kualitas tembakau. Dengan metode featureextraction pada gambar dengan menggunakan kamera maka, sebelum dilakukannya klasifikasi grade daun tembakau, pada penelitian ini melakukan tahap image acquitition, image pre-processing dan featureextraction untuk mendapatkan dataset berupa nilai angka yang selanjutnya dilakukan tahapan klasifikasi. Berdasarkan hasil pengujian yang dilakukan telah ditunjukkan bahwa sistem telah mampu mebedakan kualitas dari daun sampel tembakau dengan metode konversi RGB image to HSV Thresholding.
There are some researches done on automated segmentation of the optic nerve head [4-12, 15-26]. The optic disc and cup regions are segmented from monocular retinal images by an automatic optic disc parameterization technique is proposed by Gopal et al. Robustness against variations in an image in optic disc segmentation is achieved by using the local image information in multidimensional feature space . Kavitha et al. proposed two methods to extract the optic disc automatically i.e. using the component analysis method and region of interest based segmentation .
Fix location and size sub-block is commonly used in Region Based Image Retrieval. It divides an image into several regions or sub-blocks 𝑏 × 𝑏 in a certain size. Fix location and size sub-block selects the sub-block based on the region of image. If there are several detected objects in all regions of image, sub-blocks will be made in all regions. These sub-blocks may be irrelevant because all re- gions of image will be used. Fix location and size sub-block also has a weakness in finding the re- levant sub-block in image containing small object. It occurs because the size of an object does not meet the minimum threshold to become the rele- vant sub-block. Another reason is because the use of fix location sub-block. Fix location sub-block cannot adapt the location of an object which means that it can slice and divide one object into several objects if the location or size of the object exceeds the region. If the size of a divided object does not meet the minimum threshold, its region will not be selected as a sub-block even if the actual size of an object meets the minimum threshold. In addition, although method  uses the flexible region, that region has to be equal or greather than the threshold to be the ROI. It makes this method not efficient if there is no ROI in the image because the method will use the entire image to campare. So, the fle- xible sub-block which is able to adapt the size and the location based on the detected object is needed. In this paper, we propose a new method for local featureextraction by determining the flexible size and location of sub-block based on the tran- sition region in region based image retrieval. Ex- tracted local feature from the flexible sub-block can satisfy the user because it can handle any size of the detected object in the query image and im-