Table of ContentsINTRODUCTIONPROBLEM STATEMENTPROPOSED SYSTEM IMPLEMENTATION METHODOLOGYModule DescriptionSkin Color Segmentation:Face Detection:Eye Detection:Lips Detection:Apply the Bézier curve on lips:Applying Bézier curve on eyes:Database and coachingEmotion detection:CONCLUSIONSAbstract —Facial emotion recognition (FER) is an important topic in the field of computer vision. Actions, postures, facial expressions and language; these are considered channels for transmitting human feelings. Extensive research has been conducted to study the connections between these channels and these feelings. This paper proposes a framework that consequently perceives the feeling depicted on the human face. The results of the neural networks are combined with the results of image processing which are used to organize common emotions: Joy, ambiguous. Shaded front side images are provided as a contribution to the frame. Once the face is distinguished, an image management-based component point extraction strategy is used to separate a set of chosen highlight foci. Finally, a set of qualities acquired as a result of the preparation of these separate elements indicates that the neural system can perceive the contained feeling. Say no to plagiarism. Get a tailor-made essay on “Why Violent Video Games Should Not Be Banned”? Get the original essay Index Terms: Emotions, Feature Extraction, Neural Network, Emotion Recognition, IRON.INTRODUCTIONFeelings are human behavioral emotions that play a daily role in our daily lives in activities, e.g. leadership basics, learning, inspiration, reflection, mindfulness, organization and many more. People can identify faces and decipher passionate articulations without any real problem. Given the growing enthusiasm for sentiment recognition using apps, if robot systems effectively perceive human feelings, it paves the way for robot investigation of human behavior full of feelings and attracts attention of specialists from different fields, for example brain science, semantics, software engineering and other related orders also open up a major field of research. PROBLEM STATEMENT Human emotions and intentions are expressed via facial expressions and deriving a powerful function is the essential aspect of the facial expression system. Automated recognition of facial expressions can be an important element of natural human-machine interfaces; it can also be used in behavioral technology and medical practice. An automated facial expression popularity machine aims to solve the following problems: face detection and region in a cluttered scene, facial feature extraction, and facial feature classification. PROPOSED SYSTEM In this work, executives who will productively perceive the overall feeling of joy and ambiguity of 2D colorful face images. The work has been limited to general feelings because ordering and identifying other superficial feelings is tricky. The framework can be broadly organized into three phases (Fig.1): Face Detection Feature Extraction Facial Expression Classification Two face identification calculations are performed for face area assurance arrangement. Eyes, mouth and eyebrows are distinguished as basic highlights and their indicating components are extracted to identify the feeling. These homes ofcomponents are separated from the chosen inclusion zones using a corner point discovery calculation. Once the elements are extracted, a neural system approach is used to perceive the feelings contained inside the face. IMPLEMENTATION METHODOLOGY The overall system was developed using ASP.NET with C#. The flowchart of the system modules is shown in Fig.2 Description of modules Skin color segmentation: If the largest connected region has the probability of becoming a face, it will open a replacement type with the largest region connected. If the height of the largest connected region is greater than or equal to fifty and the height/width relationship is between one and two, then it must be heads. We then need to look at becoming the face of the largest connected region. For color segmentation, we tend to differentiate the image first. Next, we need to find the largest connected region. Then we tend to perform color segmentation (Fig.3). Face Detection: For face detection, we first tend to convert the binary image from the RGB image. Then we tend to try to notice the edge from the binary image. We tend to start scanning from the center of the image, then look for an eternal white element, then an eternal black pixel. Next, we want to find the largest dimension of the white element by looking vertically at each left and right side. Then, if the new dimension is less than half of most of the previous dimensions, then we tend to stop the scan because if we tend to reach the eyebrow, this example can occur. Then we tend to cut the face from the starting position of the forehead and its height will be multiplied by 1.5 times its dimension. We will then have an image that can contain only the eyes, nose and lips. Next, we will slice the RGB image based on the binary image. (Fig.5) Eye detection: For eye detection, we tend to convert RGB face to binary face. Now we tend to look at the dimension of the face by W. we tend to scan from W/4 to (WW/4) to find the central position of the 2 eyes. The best white continuous element on the tip between the rows is the middle position of the 2 eyes. Then, we discover the initial high or higher position of the 2 eyebrows by looking vertically. For the left eye we tend to look for w/8 to the middle and for the right eye we tend to look for the middle to w – w/8. Here, w is that the dimension of the image and the middle is that of the median position of the 2 eyes. There are also a few white pixels between the eyebrow and the eye. To form the connected eyebrow and eye, we tend to place continuous black pixels vertically from the forehead to attention. For the left eye, the vertical black pixel lines are placed between mid/2 and mid/4 and for the right eye, the lines are between mid+(w-mid)/four and mid+3*(w-mid )/four. and the height of the black pixel lines are from the eyebrow start height to (eyebrow start position h)/4. Here w is the dimension of the image and the middle is the middle position of the 2 eyes and h is the height of the image. Then we discover the low position of the 2 eyes by looking at the black element vertically. For the left eye, we tend to look for size mid/4 to mid-mid/4. And for the right eye, we tend to look for mid+(w-mid)/ four to mid+3*(w- mid)/ four dimensions from the bottom end of the image to the starting position of the eyebrow. Next we find out the correct facet of the eyeright. the left eye looking at the black element horizontally from the middle position to the start position of the black pixels between the upper position and the lower position of the left eye. And on the left facet of the right eye, we tend to look for the middle position to the start of the black pixels between the top position and the bottom position of the right eye. The left facet of the left eye is the initial dimension of the image and the right facet of the right eye is the final dimension of the image. Then we tend to cut the top position, bottom position, left facet as well as right facet of the 2 eyes from the RGB image. Lip detection: For lip detection, we tend to check the lip box. and that we envision the lip being inside the lip box. So first we tend to check the space between the forehead and the eyes. Then we tend to add the space with the lowest height of attention to see the highest height of the box that is capable of containing the rim. Now the goal {start line/start location} of the box will be the ¼ position of the left eye box and the end point will be the ¾ position of the correct eye box. and the finished height of the box will also be the bottom end of the face image. So this box can only contain the lip and part of the nose. Next, we will cut the RGB image according to the box. So for eye and lip detection, we tend to convert only the binary image from the RGB image and some looking among the binary image. Apply the Bézier curve to the lip: In the lip box, there is the lip and there will be part of the nose. So around the box there is color or skin. So we tend to convert the skin element to white element and the different elements to black. We tend to jointly notice things that are the same as skin pixels and convert them to white pixels. Here, if the distinction of RGB values ​​of 2 pixels is less than or equal to ten, then we tend to call them similar elements. Here we tend to use a bar chart to locate the space between the lower average RGB value and the better average RGB value. If the space is less than seventy then we tend to use seven to locate a similar item and if the space is greater than or equal to seventy then we tend to use ten to locate a similar item . So the value of locating a similar element depends on the quality of the image. If the image quality is high we tend to use seven to locate a similar item and if the image quality is low we use 10. So in binary image there are black areas on the lip, nose and another very small half. have a little totally different color. Then we tend to apply a huge connected region to locate the black region which contains the lip in the binary image. and that we are sure that the big connected region is the lip, because inside the lip box, the lip is the biggest problem which is totally different from the skin. Next, we need to use the Bézier curve on the binary lip. To apply the Bézier curve, we discover the start and end element of the lip horizontally. Then we tend to draw 2 tangents on the upper lip from the start and end element and jointly notice 2 points on the tangent which are not part of the lip. For the lower lip, there are 2 methods similar to those for the upper lip. we tend to use Bézier curves..