針對嵌入式機器視覺系統向獨立化、智能化發展的要求，介紹了一種嵌入式視覺系統--智能相機。基于對智能相機體系結構、組成模塊和圖像采集、傳輸和處理技術的分析，對國內外的幾款智能相機進行比較。綜合技術發展現狀，提出基于FPGA+DSP模式的硬件平臺，并提出智能相機的發展方向。分析結果表明，該系統設計可以實現脫離PC運行，完成圖像獲取與分析，并作出相應輸出。
Abstract:
This paper introduced an embedded vision system-intelligent camera ，which was for embedded machine vision systems to an independent and intelligent development requirements. Intelligent camera architecture， component modules and image acquisition， transmission and processing technology were analyzed. After comparing integrated technology development of several intelligent cameras at home and abroad， the paper proposed the hardware platform based on FPGA+DSP models and made clear direction of development of intelligent cameras. On the analysis of the design， the results indicate that the system can run from the PC independently to complete the image acquisition and analysis and give a corresponding output.
使用Nios II軟件構建工具
This chapter describes the Nios® II Software Build Tools (SBT), a set of utilities and
scripts that creates and builds embedded C/C++ application projects, user library
projects, and board support packages (BSPs). The Nios II SBT supports a repeatable,
scriptable, and archivable process for creating your software product.
You can invoke the Nios II SBT through either of the following user interfaces:
■ The Eclipse™ GUI
■ The Nios II Command Shell
The purpose of this chapter is to make you familiar with the internal functionality of
the Nios II SBT, independent of the user interface employed.
針對嵌入式機器視覺系統向獨立化、智能化發展的要求，介紹了一種嵌入式視覺系統--智能相機。基于對智能相機體系結構、組成模塊和圖像采集、傳輸和處理技術的分析，對國內外的幾款智能相機進行比較。綜合技術發展現狀，提出基于FPGA+DSP模式的硬件平臺，并提出智能相機的發展方向。分析結果表明，該系統設計可以實現脫離PC運行，完成圖像獲取與分析，并作出相應輸出。
Abstract:
This paper introduced an embedded vision system-intelligent camera ，which was for embedded machine vision systems to an independent and intelligent development requirements. Intelligent camera architecture， component modules and image acquisition， transmission and processing technology were analyzed. After comparing integrated technology development of several intelligent cameras at home and abroad， the paper proposed the hardware platform based on FPGA+DSP models and made clear direction of development of intelligent cameras. On the analysis of the design， the results indicate that the system can run from the PC independently to complete the image acquisition and analysis and give a corresponding output.
基于通用集成運算放大器，利用MASON公式設計了一個多功能二階通用濾波器，能同時或分別實現低通、高通和帶通濾波，也能設計成一個正交振蕩器。電路的極點頻率和品質因數能夠獨立、精確地調節。電路使用4個集成運放、2個電容和11個電阻，所有集成運放的反相端虛地。利用計算機仿真電路的通用濾波功能、極點頻率和品質因數的獨立控制和正交正弦振蕩，從而證明該濾波器正確有效。
Abstract:
A new multifunctional second-order filter based on OPs was presented by MASON formula. Functions， such as high-pass， band-pass， low-pass filtering， can be realized respectively and simultaneously， and can become a quadrature oscillator by modifying resistance ratio. Its pole angular frequency and quality factor can be tuned accurately and independently. The circuit presented contains four OPs， two capacitors， and eleven resistances， and inverting input of all OPs is virtual ground. Its general filtering， the independent control of pole frequency and quality factor and quadrature sinusoidal oscillation were simulated by computer， and the result shows that the presented circuit is valid and effective.
Contains a complete archiver by Haruhiko Okumura. The archiver uses
an LZ engine whose output is compressed with static Huffman. The
file is chopped into pieces which all get their own independent
set of Huffman tree s. ARJ 2.41, ZIP 2.0 and PKZIP 2.04g use
essentially the same method. ZOO and LHA even use excactly the
same method.
The GRLIB IP Library is an integrated set of reusable IP cores, designed for system-on-chip (SOC) development. The IP cores are centered around the common on-chip bus, and use a coherent method for simulation and synthesis. The library is vendor independent, with support for different CAD tools and target technologies. A unique plug&play method is used to configure and connect the IP cores without the need to modify any global resources.
ICA can be used in brain activation studies to reduce the number of dimension and filter out independent and interesting activations. This demonstration shows two studies. One provided by Hvidovre Universitets Hospital, Denmark, that consists of fMRI scannings of humans. Another provided by the EU sponsored MAPAWAMO project from fMRI scannings of monkeys. In the demo comparison between icaMS, icaML, icaMF, icaMF (positive sources) and PCA can be made. More detailes can found in [2].
The Molgedey and Schuster decorrelation algorithm, having square mixing matrix and no noise . Truncation is used for the time shifted matrix, and it is forced to be symmetric . The delay Tau is estimated .
The number of independent components are calculated using Bayes Information Criterion (BIC), with PCA for dimension reduction.
Although there has been a lot of AVL tree libraries available now, nearly all of them are meant to work in the random access memory(RAM). Some of them do provide some mechanism for dumping the whole tree into a file and loading it back to the memory in order to make data in that tree persistent. It serves well when there s just small amount of data. When the tree is somewhat bigger, the dumping/loading process could take a lengthy time and makes your mission-critical program less efficient. How about an AVL tree that can directly use the disk for data storage ? If there s something like that, we won t need to read through the whole tree in order to pick up just a little bit imformation(a node), but read only the sectors that are neccssary for locating a certain node and the sectors in which that node lies. This is my initial motivation for writing a storage-media independent AVL Tree. However, as you step forth, you would find that it not only works fine with disks but also fine with memorys, too.
很好的linux內核調試軟件 兼轅馬，沒有密碼。
The ia64 and ix86 directories contain versions of kdb prior to v2.0
(kdb version v2.0, not the kernel version). Older versions of kdb had
complete patches for each architecture it supported, each patch included
all the common kdb code. This format was awkward to maintain and use
for multiple platforms.
Starting with kdb v2.0 there is a common patch against each kernel which
contains all the architecture independent code plus separate architecture
dependent patches. Either use an old style (v1.8 or v1.9) kdb patch or
use a new style (v2.0) common patch plus the corresponding architecture
dependent patch.
