Human face recognition is one of the most challenging topics in the areas of image processing, computer vision, and pattern recognition. Before recognizing the human face, it is necessary to detect a face then extract the face features. Many methods have been created and developed in order to perform face detection and two of the most popular methods are Viola-Jones Haar Cascade Classifier (V-J) and Histogram of Oriented Gradients (HOG). This paper proposed a comparison between VJ and HOG for detecting the face. V-J method calculate Integral Image through Haar-like feature with AdaBoost process to make a robust cascade classifier, HOG compute the classifier for each image in and scale of the image, applied the sliding windows, extracted HOG descriptor at each window and applied the classifier, if the classifier detected an object with enough probability that resembles a face, the classifier recording the bounding box of the window and applied non-maximum suppression to make the accuracy increased. The experimental results show that the system successfully detected face based on the determined algorithm. That is mean the application using computer vision can detect face and compare the results.

Image reconstruction of traffic signs serves in order to improve the image quality of traffic signs that are covered leaves, tree branches or pole so that it can facilitate the process of recognize or detecting traffic signs. Inpainting is one of the image restoration methods to reconstruct images that have been damaged or the removal of unwanted objects where the area to be restored is based on information around the area. Criminisi introduces an inpainting method with an exemplar-based approach that combines the structure-oriented and texture-oriented scheme. While Hung popularized the Criminisi method and combined it with contour construction (Bezier curve). The researcher tried to compare the two methods by using a number of images of traffic signs, one of them by using images with a size of 400 x 407 pixels, issuing the results of Peak Signal to Noise Ratio (PSNR), the Criminisi method gets results from PSNR 36,479 decibel (dB), while the Hung method gets results from PSNR 36,827 dB. This means the result of Hung method is more similar to original image than Criminisi method.

The coin recognition system is a very important role in currency processing systems which is aims to minimize human miscalculation and time to calculate coin values. The coin identifier and recognition to calculate the value of Indonesian coins based on android developer presented in this paper. The two coin parameters are used for coin recognition which are radius and colour of the coins. The Circle Hough Transform (CHT) method with gray scaling and erosion process as pre-processing image used to calculate radius parameters, based on android developer and computer vision. After the radius value is obtained then the average RGB value of the coins is detected to determine the colour parameters. By calculating the diameter and determining the colour, the coins can be recognized and the total value of the coins in the input image can be known. The experimental results show that the system has been obtained indicate that the system successfully recognized coins based on determined algorithm. That is mean the application using computer vision can recognition coin and count the value of coins.

Sending secret message need an algorithm that can protect message to receiver without knowing by other and keep the originality of message. Previous research, encoding secret message in video inserted in frame while audio secret message inserted in bit audio. it makes it easier for someone to detect the presence of a message. In this paper, try to encode the secret message at video file to encode in audio. Audio separated from video in order to encode secret message in audio. Echo Hiding method used in encode secret message to audio process. Testing use Bit Error Rate (BER) method to analysis modification in audio and Normalized Correlation (NC) analysis authentic the secret message. Based on the experiment, the results obtained were NC values (0.98-1) and BER values (0-3.12). based on the results obtained by the method we can insert messages better than the previous method with the addition of noise elimination in audio.