No description available.

Quantum Well Infrared Photodetector (QWIP) technology is promising for the development of large format low cost single and dual/multi color infrared sensor arrays. Thanks to the mature III-V semiconductor technology, QWIP focal plane arrays (FPAs) provide high uniformity and excellent noise equivalent temperature difference (NETD) in both long wavelength infrared (LWIR 8-12 &amp / #61549 / m) and mid wavelength infrared (MWIR 3-5 &amp / #61549 / m) bands. This thesis work focuses on the development of large format single and dual color MWIR QWIP FPAs. For single band MWIR detection, we report QWIP FPAs on InP substrate as an alternative to the GaAs based MWIR QWIPs suffering from the degrading effects of lattice mismatched epitaxy. In the course of this work, epitaxial growth conditions of the device structure were optimized and 640&times / 512 AlInAs/InGaAs QWIP FPAs on InP substrate have been fabricated yielding NETD of 22 mK (f/1.5) and background limited performance (BLIP) temperature as high as 115 K In the second part, we report the first voltage tunable 640&times / 512 dual color MWIR QWIP FPA. After optimizing epitaxial growth of AlGaAs/InGaAs material system, we have designed and implemented the device structure to yield voltage tunable spectral response in two different windows in the MWIR band. The FPA provides NETDs of 60 and 30 mK (f/1.5) in colors 1 and 2. The results are very encouraging for the development of low cost dual/multi color FPAs since our approach utilizes one In bump per pixel allowing fabrication of dual color FPAs with the same process steps for single color FPAs.

The diploma thesis focuses on programming in the Matlab for the Raspberry Pi 3 platform. For the purpose of the presentation, there are two applications designed for Raspberry Pi that are using available hardware, camera and servos. The first application serves as colour object detecting and accurate tracking by using camera calibration. The second application serves as a face detection and recognition. These applications are implemented by modern methods and knowledge of computer vision. Tracking of the objects and face recognition are verified by an experiment that reveals the accuracy of the used methods.

在不同的光源環境下，顏色的資訊會隨著環境而改變，因此要做到穩定的膚色辨識並不容易。先前的研究指出，人類的膚色大致上可歸納於特定顏色座標系統中的一段範圍內。但是根據我們的實驗，這段範圍會隨著環境光源的改變而產生偏移，因此運用相同的標準，無法在不同的場景下做到準確的膚色辨識與切割。針對這個議題，我們提出利用影像中非色彩的資訊，找出最符合膚色的範圍。具體來說，借重臉部偵測獨立於色度的特性，我們使用臉部偵測的結果，做為調整膚色範圍的依據。實驗所得到的辨識結果，效能與效率都足以運用在以視覺為基礎的人機介面，例如手部與指尖偵測。 / Robust detection of skin color is a difficult task since color information changes under different lighting conditions. Previous research indicated that human skin color is restricted to a small range of values in certain color coordinate systems. However, the ranges tend to shift with varying illumination according to our experiments. It is therefore unattainable to derive a universal standard for skin color detection and segmentation for general scenes. To address this issue, we propose to use achromatic features to identify the proper ranges of skin color in an image. Specifically, we utilize the result of face detection, which is independent of chromatic properties, to guide the process of skin color range selection. Experimental results have validated the efficacy and efficiency of the proposed methodology for vision-based human-computer interface such as hand and finger detection.

The aim of the thesis is detection of a known object in series of pictures. Detection is divided into two methods. First method is based on edge and color detection and comparison. Edge detection is based on detection using both Gradient and Laplacian, so on the first-order and the second-order derivative. Sobel operators were used as well as Laplacian of gaussian method. Thresholding is also used as well as autothreshold calculation. There are two variants of color detection considered in the thesis, direct color comparison and detection based on interest color search. The second part of the thesis is based on interested point detection using a modified SURF method to detect a known object in series of pictures.

This master’s thesis is dealing with recognition of an easy static gestures in order to computer controlling. First part of this work is attended to the theoretical review of methods used to hand segmentation from the image. Next methods for hang gesture classification are described. The second part of this work is devoted to choice of suitable method for hand segmentation based on skin color and movement. Methods for hand gesture classification are described in next part. Last part of this work is devoted to description of proposed system.