Hasil Pencarian  ::  Simpan CSV :: Kembali

"Pada penelitian ini akan diterapkan sistim visi komputer terhadap contoh maket pemandangan suatu daerah, yang terdiri dari beberapa jenis kelas pola, misalnya tanaman, air, perumahan, dan sebagainya. Data citra pemandangan tersebut direkam oleh suatu sistim akuisisi memakai kamera video CCD-warna yang mengandung informasi digital dalam tiga kanal spektrum elektromagnetik.Yang akan diselidiki adalah bagaimana jenis-jenis pola tersebut dapat diklasifikasikan oleh suatu sistim perangkat komputer cerdas berbantuan jaringan syaraf tiruan dan logika fuzzy, sehingga hasilnya dapat tervisualisasi ''serta memberikan unjuk- kerja klasifikasi yang cukup .memadai dibandingkan dengan metode-metode yang telah lazim digunakan, seperti multiple density slicing, nearest neighbor, dan maximum likelihood.Aspek penelitian.ini adalah bahwa kalau sistim tersebut berhasil, maka baik metode maupum perangkat yang dibuat dapat dikembangkan untuk teknik penginderaan jauh, aplikasi medis, kontrol kualitas dengan.pemeriksaan oleh mesin komputer, dsb.Unjuk kerja metode klasifikasi dinyatakan oleh prosentasi kebenaran pada-suatu tabel yang menyatakan distribusi pengkategorian obyek ke dalam kelas yang telah ditentukan sebelumnya. Pengecekkannya langsung dilakukan dengan maket yang dibuat, sehingga beberapa pengamatan lapangan dari berbagai sudut pandang serta ketinggian dapat diatur sebaik-baiknya guna melengkapi hasil-hasil percobaan. Disamping itu hasil klasifikasi yang bertipe peta tematik disertai legends yang sesuai, dapat ditampilkan atau divisualisasikan pada layar monitor SVGA.

---

In this research, it is applied a computer vision system to an image which is consisting of several objects patterns. from an artificial maquette scene which had been taken by a color CCD camera. Due to object's responses in several electromagnetic waves are different to each other, then the 'recorded image can be splitted into three different color channels, i.e. blue, green, and red.

The research will investigate how to classify the above patterns. by using' an. intelligent computer system such as neural networks and fuzzy logic in order to obtain a reasonable performance compared with the available conventional classification system such as multiple density slicing, nearest. neighbor, and maximum likelihood.

The aspect of research is that the designed method -if successful- may be developed-to be applied to remote sensing technology, medical application, quality control by machine inspection, etc.

The classification performance is represented by percentage of correct on a truth table, which is reflected the distribution of object's category to a predetermined category. The direct observation can easily be done on the available maquette, so the several looking angles and height can be arranged to accomplish the experimental results. Beside it, the classification results will be represented on a thematic map with suitable legends to be visualized on a SVGA color monitor."

"ABSTRAKPada penelitian tahapan pertama ini telah dilakukan pengambilan data dengan menggunakan kamera CCD (Charge Coupled Device) berwarna terhadap suatu maket pemandangan yang terdiri dari beberapa kelas obyek karena respon sesuatu obyek terhadap beberapa panjang gelombang elektromagnetik adalah berbeda-beda, maka data hasil rekaman oleh kamera tersebut dipecah menjadi tiga jenis warna yaitu biru, hijau, dan merah.Untuk melaksanakan pemisahan kanal panjang gelombang elektromaknetik tersebut, telah dibuat suatu perangkat lunak komputer ImagePro, yang ditulis dalam bahasa pemrograman C, dan bekerja di bawah MS-Windows 3.1. yang dapat dijalankan pada IBM/PC-AT microcomputer jenis 286 ke-atas. Beberapa fasilitas tambahan seperti : image enhancement, filtering, cutting, statistical properties, format exchange, dual and frame processing, dan sebagainya telah pula diikutsertakan.Proses klasifikasi tanpa disupervisi atau clustering diterapkan dengan bantuan topologi ART2 dari jaringan Syaraf Tiruan untuk melihat bagaimana informasi data obyek tersebut dapat diklasifikasikan menjadi beberapa kelas yakni air, jalanan, tumbuhan, rumah atap genteng, dan rumah atap beton.Hasil klasifikasi dengan memanfaatkan jaringan syaraf buatan tersebut untuk tahap pertama ini telah cukup memuaskan secara kualitatif; walaupun hasil tersebut harus diuji tingkat ketelitiannya pada tahapan berikutnya dari rangkaian penelitian yang tengah dilaksanakan ini.

---

ABSTRACTIn the first stage of this research, it is done the data acquisition to several classes of objects on an artificial maquette scene by using a color CCD camera Due to object's responses in several electromagnetic waves are different to each other, then the recorded data can be splitted into three different colors channels, i.e. blue, green, and reed.In order to split those electromagnetic channels, it is designed a computer software called ImagePro, which is written in C-language, under MS-Windows 3.1, and can be run on an IBMIPC-AT microcomputer 286 processor or above. Several additional supporting features such as : image enhancement, filtering, cutting, statistical properties, format exchange, dual and frame processing, etc. are also included in the software.The unsupervised classification by using an artificial neural networks ART2 topology is applied, to observe how the those object's data can be classified. into several classes. i.e.: water, road, vegetation, red roof and concrete roof houses.The result of classification at the initial stage of the research provides an acceptable qualitative performance, although its accuracy should be tested later at the need stage of this midyears research plan."

"ABSTRAKPada penelitian tahapan kedua ini telah dilakukan koreksi geometrik terhadap citra hasil pengambilan (akuisisi) data yang dikerjakan pada tahap tahap pertama dengan menggunakan kamera CCD (Charge Coupled Device) berwarna terhadap suatu maket pemandangan yang terdiri dari beberapa kelas obyek. Karena respon sesuatu obyek terhadap beberapa panjang gelombang elektromagnetik adalah berbeda-beda, maka data hasil rekaman oleh kamera tersebut dipecah menjadi tiga jenis warna yakni : biru, hijau, dan merah., sedangkan pada penelitian untuk koreksi ini hanya dipakai kanal biru.Untuk melaksanakan koreksi geometrik tersebut, telah dibuat suatu perangkat lunak komputer GeoPro, yang ditulis dalam bahasa pemrograman C, dan bekerja di bawah MS-DOS serta dapat dijalankan pada IBMIPC-AT microcomputer jenis 386 ke-atas. Beberapa fasilitas tambahan seperti : histogram display, 3-D display, mozaicking, rotating, trimming, flipping, dan sebagainya telah pula-diikut sertakan.Proses koreksi ini diterapkan guna menguji unjuk kerja beberapa pasangan teknik transformasi dan teknik resampling, terutama waktu proses serta hasil citra keluaran. Dari hasil yang diperoleh ternyata bahwa untuk citra masukan yang digunakan, maka metode transforrnasi dengan memakai orde satu dan teknik resampling dengan cara nearest-neigbor telah cukup memadai dengan proses eksekusinya yang relatif lebih cepat.

---

ABSTRACT

In the second stage of this research, it is done the geometric data correction to an image consisting of several objects from an artificial maquette scene which had been taken by a color CCD camera. Due to object's responses in several electromagnetic waves are different to each other, then the recorded data can be splitted into three different colors channels, i.e. blue, green, and red, but in this correction research it is only used the blue channel.

In order to conduct the above correction, it is designed a computer software called GeoPro, which is written in C-language, under MS-DOS, and can be run on an IBMIPC-AT microcomputer 386 processor or above. Several additional supporting features such as : histogram display, 3-D display, mozaicking, rotating, trimming, flipping,, etc. are also included in the software.

The correction is applied for testing the performance of several combination of transformation methods and resampling methods, especially for computation time, and also results of the corrected images visually. From the obtained results, it is concluded that the combination of first order transformation method with the nearest neighbor resampling method, has provided an adequate result due to its faster excecution time and the acceptable corrected images."

"In this research, it is designed a simple inspection model for defect detection on woven fabrics at the weaving stage of processing, based on texture analysis. Textural features that extracted by using the NGLDM (Neighboring Greylevel Dependence Matrices) from the several avail-able samples either for normal or defective weaving products, are intelligently recognized by a neural network computational system. The model is useful in textile industry, which provide woven qualities produced by weaving machines, therefore, from the defect's information one can separates those products which have different grades to be processed at the dye finishing stage, and may check previous yarn's treatment, mechanical failures, etc. The inspection system is equipped by a flatbed conveyor, a CCD camera, and a microcomputer IBM-PC/AT 386 with a Computer Eyes image grabber card. The testing results of defect detection on the available samples, indicate more than 80% of recognition level can be achieved. In the future, it is anticipated that the system may be developed, in order to reduce much more human intervention for the defect detection."

"In this research it is applied a pattern recognition system by using an artificial neural networks to recognize several samples on weaving products, such as plain weave, twill weave, and sateen weave. In order to extract textural characteristics or features from sample images, it is used the Neighboring Grey Level Dependence Matrix (NGLDM)-method as proposed by Sun, which is invariant under rotation and linear grey level transformation. Five textural features i.e. Small Number Emphasis, Large Number Emphasis, Number Non uniformity, Second Moment, and Entropy will be used as the representative features of sample images. Those features are used as input to the neural networks, which have learned by the back propagation method. Baths methods (continuous and periodic) for changing the interconnection weights, and the performances of the two types of neuron transfer functions are also observed and investigated, in order to obtain an optimal network configurations. The results of experiment will be very useful for the next stage of research in designing an integrated vision system for the recognition of weaving product's quality in textile industry."