14 research outputs found
COMPARISON OF IMAGE SEGMENTATION METHOD IN IMAGE CHARACTER EXTRACTION PREPROCESSING USING OPTICAL CHARACTER RECOGINITON
Full text linkToday, there are many documents in the form of digital images obtained from various sources which must be able to be processed by a computer automatically. One of the document image processing is text feature extraction using OCR (Optical Character Recognition) technology. However, in many cases OCR technology are unable to read text characters in digital images accurately. This could be due to several factor such as poor image quality or noise. In order to get accurate result, the image must be in a good quality, so that digital image need to be preprocessed. The image preprocessing method used in this study are Otsu Thressholding Binarization, Niblack, and Sauvola methods. While the OCR technology used to extract the character is Tesseract library in Python. The test results show that direct text extraction from the original image gives better results with a character match rate average of 77.27%. Meanwhile, the match rate using the Otsu Thressholding method was 70.27%, the Sauvola method was 69.67%, and the Niblack method was only 35.72%. However, in some cases in this research the Sauvola and Otsu methods give better results
Perbaikan Citra Tanda Tangan Digital Menggunakan Metode Otsu Thressholding dan Sauvola
Full text linkAbstract : The digital era forces people to digitize in all fields. Digital product that is often found is digital image. One kind of digital image application is the use of a digital image signature embedded in a document. However, often the results are unsatisfactory, such as background color problems, noise, lack of clarity, etc. The quality of digital image signatures can be improved by implementing the Otsu Thresholding and Sauvola methods. These two methods were chosen because they are widely used in document image quality improvement. The purpose of this study is to produce a better digital image signature and to compare the performance of these two methods. The results showed that the quality of the images produced by these two methods was better than the original image or by using standard filters from a word processing application. Meanwhile, from the two methods used, Sauvola\u27s method was slightly better than Otsu\u27s method. In terms of visual evaluations, the Sauvola method total score is 13 compare to Otsu method score that is 11. Meanwhile, the PSNR ratio show that the two methods give the same results, that is 34,571 db
FORECASTING HARGA SAHAM MENGGUNAKAN METODE SIMPLE MOVING AVERAGE DAN WEB SCRAPPING
Full text linkAbstract: The fluctuative of stock prices in a secondary market provide the possibility for investors/traders to gain profits through the difference in stock prices (capital gain). In order to obtain these benefits, it is necessary to analyze before buying shares, through fundamental and technical analysis. One of several methods in Technical Analysis is Simple Moving Average Method. This method can be used to predict (forecast) stock prices by calculating moving average of the stock price history. Historical stock prices can be obtained in real time using the Web Scrapper technique, so the results is more quickly and accurately. Using the MAPE (Mean Absolute Percent Error) method, the level of accuracy of forecasting can be calculated. As a result, the program was able to run successfully and was able to display the value of forecasting and the level of accuracy for the entire data tested in LQ45. Besides forecasting with a value of N = 5 has the highest level of accuracy that reaches 97,6 % while the lowest one is using the value of N = 30 which is 95,0 %
Peramalan Harga Saham Menggunakan Metode Autoregressive Dan Web Scrapping Pada Indeks Saham Lq45 Dengan Python
Full text linkThe Stock Exchange gives investors or traders the possibility to gain a profit (capital gains) or losses (capital loss) due to stock prices fluctuation. This uncertainty can be circumvented by applying forecasting methods to predict future stock prices. One of the method is Autoregressive. This method uses stock data in the past to get a formula to predict future stock prices. The stock price data history can be seen at several stock data provider pages and can be retrieved automatically using the Web Scrapper technique. This tehcnique make the result can be obtained quickly, easily, and accurately. The forecasting accuracy is measured using the MAPE (Mean Absolute Percent Error) method. This method was chosen because it is easier for commoner to understand. As a result, forecasting program are succed to give stock price predictions and their accuracy. The data tested in this study are all stocks incorporated in the LQ45 index. The average accuracy level obtained was 94,62%. The highest accuracy level is BKSL stock of 99,92% and the smallest one is ASRI stock of 90.13%.
Bursa Saham memberikan kemungkinan investor untuk memperoleh keuntungan (capital gain) atau mengalami kerugian (capital loss) dikarenakan harga saham yang berfluktuasi. Ketidakpastian ini bisa disiasati dengan menerapkan metode peramalan untuk memprediksi harga saham di masa datang. Salah satu metode peramalan yang dapat digunakan adalah Autoregressive. Metode ini memanfaatkan data saham di masa lalu untuk mendapatkan formula prediksi di masa datang. History harga saham bisa dilihat secara realtime melalui beberapa laman penyedia data saham. Data ini bisa ditarik secara otomatis dengan menggunakan teknik Web Scrapper, sehingga hasil peramalan dapat diperoleh dengan lebih cepat, mudah, dan akurat. Tingkat akurasi peramalan diukur dengan menggunakan metode MAPE (Mean Absolute Percent Error). Metode ini dipilih karena lebih mudah dipahami oleh para pengguna awam. Hasilnya, aplikasi peramalan mampu menampilkan prediksi harga saham beserta tingkat akurasinya. Data yang diujikan pada penelitian adalah semua data saham LQ45. Tingkat akurasi rata-rata yang diperoleh adalah sebesar 94,62 %. Tingkat akurasi terbesar terdapat pada emiten BKSL dengan nilai persentase 99,92 % dan tingkat akurasi terkecil terdapat pada emiten ASRI dengan nilai persentase 90,13 %.
 
Analisis Sentimen Vaksinasi Booster Covid-19 pada Platform Twitter Menggunakan Metode Naïve Bayes
Full text linkSince the end of 2019, the Covid-19 virus hit the whole world, including in Indonesia. One of the efforts to deal with the Covid-19 virus is vaccination. In Indonesia, the government requires people to vaccinate 3 times, that are First Vaccination, Second Vaccination, and Booster Vaccination. The public's response to the booster vaccine are varies. This study aims to reveal public sentiment towards booster vaccine activities. The research was conducted by collecting tweet data from the Twitter platform. The research was conducted by collecting data tweets from Twitter. The method used is the Naïve Bayes Classifier because the method is simple, the process is fast, and it has a fairly high level of confidence. In this method, public sentiment is divided into three, that are positive, neutral, and negative. The results showed that most people responded positively to this booster vaccine activity with a value of 56.8%, neutral as much as 39.9%, and negative as much as 3.3% with an accuracy rate of 86%
Ekstraksi Karakter Plat Nomor Kendaraan Menggunakan Google Vision dan Regex Pattern
Full text linkDi era serba AI (Artificial Intelligence) seperti saat ini, penggunaan teknologi Optical Character Recognition (OCR) adalah hal yang lumrah. Salah satunya adalah pada Manless Parking System atau sistem parkir otomatis. Pada sistem ini teknologi OCR digunakan untuk mendeteksi plat nomor kendaraan secara otomatis. Namun begitu tidak semua metode memberikan hasil yang optimal. Pada penelitian-penelitian sebelumnya rata-rata tingkat akurasi dari penerapan OCR adalah sebesar 90%. Pada penelitian ini digunakan metode OCR dari Google yaitu Google Vision dan Pattern Regex untuk mendeteksi plat nomor kendaraan pada citra digital. Google Vision dipilih karena kemampuannya yang relatif lebih baik dibandingkan metode OCR lainnya seperti Tesseract dan EasyOCR. Sedangkan Regex Pattern digunakan untuk melakukan filtering terhadap teks hasil ekstraksi OCR yang tidak diperlukan. Penelitian ini diujikan terhadap 150 data sample berformat JPG dan JPEG. Ukuran file bervariasi antara 50 KB – 2 MB. Sudut pengambilan gambar juga dibuat bervariasi. Hasil penelitian menunjukkan bahwa penggunaan metode OCR dengan Google Vision dan Regex Pattern memberikan hasil yang sangat baik dengan tingkat akurasi mencapai 99,50%. Hasil ini lebih baik dibandingkan hasil pada penelitian-penelitian sebelumnya.
Kata Kunci : Plat Nomor Kendaraan, Optical Character Recognition, Google Vision, Regex Patter
Perbaikan Citra Dokumen Hasil Pindai Menggunakan Metode Simple, Adaptive-Gaussian, dan Otsu Binarization Thresholding
Full text linkThe use of digital images from scanned documents is commonly used both for data backup and for further processing. However, often the digital image obtained is not optimum due to various factors like noise. The method to improve the quality of digital images is to filter images using the Thresholding method. This study compares three Thresholding methods, which are Simple Thresholding, Adaptive-Gaussian Thresholding, and Otsu Binarization. All three methods have advantages and disadvantages. However, using the MSE and PSNR assessment parameters, the Simple Thresholding method shows better quality with an MSE value of 5,196.76, followed by Otsu Binarization with a value of 5,934.10, and Adaptive-Gaussian Thresholding with a value of 9,025.29. Meanwhile, by using PSNR, the value for Simple Thresholding is 13.37, followed by Otsu Binarization with a value of 12.47, and Adaptive-Gaussian Thresholding with a value of 10.31
STEEL BOX GIRDER BRIDGE COMPONENT TRACEABILITY SYSTEM USING TREE STRUCTURE DIAGRAM AT PT BUKAKA TEKNIK UTAMA
Full text linkThe International Organization for Standardization (ISO) through ISO 9001:2015requires every product to have product traceability. In response to these challenges, PT Bukaka Teknik Utama developes the Traceability System in the Steel Box Girder Bridge products. Traceability System built by adopting Tree Structure Diagram Concept to describe production system process currently runs. The production process start from identify raw material, cutting process, sub-assembly process, and assembly process. This concept is then translated into Relational Database by applying Parent-Child Concept. The result of this Traceability System is the system able to issue a list of product traceability including raw material information, sub-contractor/employee who work on them, etc, quickly and accurately. System testing was carried out using the black box method, where of the 37 items tested all functioned properly. Tests were also carried out to determine the accuracy and speed of the system compared to the manual method. Of the 10 tests carried out, the system traceability is exactly the same as the manual method with an average processing time of 3 seconds, compared to the manual method, which is 97.6 seconds
PENYEBARAN DAN BUDIDAYA IKAN AIR TAWAR DI PULAU JAWA BERBASIS WEB
No full textIkan air tawar merupakan komoditas yang selama ini menjadi andalan di bidang pangan. Budidaya ikan air tawar semakin hari semakin menggiurkan. Menurut laporan Badan Pangan PBB, pada tahun 2021 konsumsi ikan perkapita penduduk dunia akan mencapai 19,6 kg per tahun. Dari sisi produksi, pada tahun 2011 produksi perikanan nasional mencapai 12,39 juta ton. Dari jumlah itu, produksi perikanan tangkap sebanyak 5,41 juta ton dan produksi perikanan budidaya 6,98 juta ton. Dari total produksi perikanan budidaya, jumlah budidaya ikan dalam kolam air tawar menyumbangkan angka hingga 1,1 juta ton. Kecilnya jumlah yang disumbangkan oleh budidaya ikan air tawar adalah karena kurangnya pengetahuan masyarakat mengenai budidaya ikan air tawar tersebut, dari mulai jenis ikan air tawar yang ada, cara budidaya dan pangsa pasar yang akan dituju. Tujuan dari penulisan ini adalah pembuatan website untuk memberikan informasi mengenai jenis ikan air tawar yang ada di pulau jawa, cara budidaya yang mudah dengan menggunakan video dan cara budidaya dengan memanfaatkan lahan yang sempit sesuai dengan ketersediaan lahan yang ada. Aplikasi berbasis Web informasi penyebaran dan budidaya ikan air tawar yang akan dibuat menggunakan dreamweaver, appserv dan phpmyadmin. Metode penelitian yang digunakan adalah dengan studi pustaka dan observasi ke tempat budidaya mengenai budidaya ikan air tawar. Kata Kunci : Aplikasi berbasis Web,budidaya, ikan air tawar, informasi, pulau jaw
KOMPARASI KECEPATAN HADOOP MAPREDUCE DAN APACHE SPARK DALAM MENGOLAH DATA TEKS
Full text linkIstilah Big Data saat ini bukanlah hal yang baru lagi. Salah satu komponen Big Data adalah jumlah data yang masif, yang membuat data tidak bisa diproses dengan cara-cara tradicional. Untuk menyelesaikan masalah ini, dikembangkanlah metode Map Reduce. Map Reduce adalah metode pengolahan data dengan memecah data menjadi bagian-bagian kecil (mapping) dan kemudian hasilnya dijadikan satu kembali (reducing). Framework Map Reduce yang banyak digunakan adalah Hadoop MapReduce dan Apache Spark. Konsep kedua framework ini sama akan tetapi berbeda dalam pengelolaan sumber data. Hadoop MapReduce menggunakan pendekatan HDFS (disk), sedangkan Apache Spark menggunakan RDD (in-memory). Penggunaan RDD pada Apache Spark membuat kinerja framework ini lebih cepat dibandingkan Hadoop MapReduce. Hal ini dibutktikan dalam penelitian ini, dimana untuk mengolah data teks yang sama, kecepatan rata-rata Apache Spark adalah 4,99 kali lebih cepat dibandingkan Hadoop MapReduce
