Guard rail的問題，透過圖書和論文來找解法和答案更準確安心。 我們查出實價登入價格、格局平面圖和買賣資訊

J3: A Ship’s Boat from the Portus Magnus, Alexandria

為了解決Guard rail的問題，作者 這樣論述：

J3, a small boat around 6 meters long, was discovered by the Institut Européen d’Archéologie Sous-Marine (IEASM) in the Eastern Harbor of Alexandria, the ancient Portus Magnus. The boat, dating to the 1-2nd century AD, was discovered towards the northeastern end of Antirhodos Island in the sheltered

inner harbor, well preserved in the sediments.The excavation allowed a team from the IEASM to study the wreck in detail, to reveal the details of the construction of this transom sterned vessel. The ship was flat-bottomed with a prominent keel, the surviving length of which was 4.4 m long. The port

side was completely preserved in the center of the boat from the keel through fourteen strakes to the guard rail. Due to the angle that it came to rest on the harbor bottom, only six lines of strakes were preserved on the starboard of J3. The boat was constructed with a well-developed internal stru

cture with 21 preserved frames. The carvel planking of the ship was assembled with mortise-and-tenon joints cut in the middle of the edges of the planks.The preservation of J3 was such that the remains of two thwarts were discovered still in situ with evidence to indicate the presence of a third row

ing bench. These were accompanied by three thole pins set into the guard rail and the final lines of strakes. The boat was preserved with a mast-step, indicating that it could be sailed as well as rowed.A wide variety of carefully chosen wood species were used in its construction and as is typical f

or the Roman period hard woods used for the keel and frames and softwoods for the planking. These species were not native to Egypt, nor were several of them used in shipbuilding in the eastern Mediterranean during the Roman period. Instead, the suite of timbers used would suggest that J3 originated

in a boatyard on the Tyrrhenian coast of Italy.J3 was transporting a large iron anchor at the moment it sank. This occupied almost the entire preserved length of the ship and was carefully positioned with its crown on the decked aft area and the shank under the removable thwarts. It had been disasse

mbled before stowage in J3 with the stock removed and carefully placed in the bottom of the boat. While typologically similar anchors have also been found in anchorages outside of the Portus Magnus, the majority of this type have been found in the western Mediterranean.

Guard rail進入發燒排行的影片

以時序錯誤導向電軌調變技術實現之細緻化電壓調節及其於能耗可調數位系統之應用

為了解決Guard rail的問題，作者陳威仁 這樣論述：

電壓調節技術(voltage scaling)在提高數位系統的能源效益方面具有相當大的潛力。然而，其節能效益在極大程度上受制於系統中穩壓電路之性能。本論文旨在提出一種可打破此限制的基於時序錯誤導向之電源軌調變技術，並以此技術實現細緻化的電壓調節。所提出之技術只需要少數電壓檔位，即可利用電源軌抖動(supply rail voltage dithering)的方式來近似出細緻化電壓調節的效果。因此，所提出之方法可以顯著降低晶片內穩壓電路的設計開銷。由於數位式低壓降線性穩壓器(digital low-dropout regulator, DLDO)具有無縫整合：（一）穩定輸出電壓、（二）電源軌抖

動、以及（三）電源閘控(power gating)等技術之特性，因此本論文利用DLDO來實現所提出之電源軌調變技術。為了精確與快速地實現適用於不同應用場景之DLDO電路，本論文也提出一種具有快速週轉時間的DLDO設計方法，並實際以一高性能DLDO設計為例驗證其效益。實驗結果指出，使用了聯電110奈米製程所製造的DLDO測試晶片展現出3毫伏特的超低漣波、67奈秒的輕載至重載暫態響應及250奈秒的重載至輕載暫態響應。與最先進的DLDO設計相比，該DLDO具有更簡潔的硬體架構且在品質因數(figure of merit)方面展現出高度競爭力。而後，本文以一種基於DLDO的抖動電源 (dithered

power supply)來實現所提出之電源軌調變技術。為了驗證所提出技術之效益，我們使用了一個具有時序錯誤偵測與修正能力之可程式化DSP資料路徑(datapath)作為測試載體。此測試晶片以台積電65奈米低功耗製程實現，而研究結果表明，所提出之電源軌調變技術有助於回收設計階段時留下之保守設計餘裕(design margin)並提高能源效率。量測結果指出，當該DSP資料路徑被程式化為一個無限脈衝響(infinite impulse response)數位濾波器以執行低通濾波時，所提技術之節能效益最高可達30.8%。最後，本論文將所提出之電源軌調變技術應用於即時影像處理系統中並探索其先天的容錯

能力。我們利用人眼視覺可將視訊中相鄰影格及影格中鄰近畫素進行視覺積分的特性，來達到即使不須對時序錯誤進行主動偵測及修正也能維持一定視覺品質的效果。因此，藉由巧妙安排容許時序錯誤發生之位置（藉由降低操作電壓），因時序錯誤所產生的錯誤畫素即可主動被人眼濾除。 該測試晶片以聯電40奈米製程實現，其搭載了一個即時視訊縮放引擎作為測試載具。在實驗結果中，該測試晶片展現了高達35%的節能效益，並能在不需對時序錯誤做出任何修正、且不須更動資料路徑架構的狀況下，仍能維持良好的主觀視覺感受。在五分制的平均主觀意見分數(mean opinion score)評量中，各類型的畫面皆達4分以上。而在客觀評量方面，峰值

信號雜訊比(peak signal-to-noise ratio)皆高於30分貝。

Ring of Stone Circles

為了解決Guard rail的問題，作者 這樣論述：

An accessible exploration of England’s prehistoric past through the clues set in stone by our ancient ancestors. Stan Abbott explores Britain’s neolithic remains, including Castlerigg and Long Meg and her Daughters.In Ring of Stone Circles, Stan Abbott sets out to explore one part of England for the

visible clues to our mysterious past from the Neolithic and Bronze Ages: stone circles and standing stones, in Cumbria--the Northern English county that boasts more of these monuments than any other. Here, the country’s tallest mountains are ringed by almost fifty circles and henges, most of them s

ited in the foothills or on outlying plateaux. But why were these built?We may never have a definitive answer to this question, but by observing and comparing sites, a greater understanding emerges. Were some circles built for ritualistic purposes, or perhaps astronomical? Were they burial sites, or

simply meeting places?Join Stan Abbott as he searches for the hidden stories these great monuments guard--and might reveal if we get to know them. Stan L Abbott is an award-winning journalist and writer on the outdoors and conservation. He has written on travel and the outdoors for national and r

egional newspapers and magazines. He has also worked in industry and tourism in the UK and Europe. He has written and edited many books about local history, travel, and walking, and published or edited several travel and conservation magazines. A prominent member of the Outdoor Writers & Photographe

rs Guild, he led a rail restoration group in Yorkshire and writes regularly for the New European newspaper.

基於語意之輪廓表示法及全連結捲積類神經網路之單晶片多車輛辨識系統

為了解決Guard rail的問題，作者徐雋航 這樣論述：

鑒於現今智慧車輛發展迅速，前方車輛辨識及車距檢測為先進駕駛輔助系統 (Advanced Driver Assistance Systems, ADAS) 設計中相當重要的一環，此項技術通常藉由攝影鏡頭擷取前方影像，並透過影像辨識技術來判斷前方是否存在車輛、障礙物等等，進而控制車輛減速以保持安全距離。而這些複雜的圖形辨識技術往往需要透過高功耗之大型運算系統來實現，並且，若將傳統電腦安裝於車內常需要克服體積過大、耐震性不佳等缺點。因此，本研究專注於如何將車輛辨識及車距檢測演算法實現於單晶片，以達到高性能、低功耗，以及體積小之目的。為實現前方車輛辨識及車距檢測，本研究透過單一彩色相機模組收集前方影

像資訊，並於單一現場可程式邏輯閘陣列 (Field Programmable Gate Array, FPGA) 晶片中以最精簡之硬體電路實現白平衡 (White Balance)、影像對比度強化技術 (Image Contrast Technique)、物體邊緣檢測、利用基於模糊語意影像描述 (Semantics-based Vague Image Representation, SVIR) 改良之基於語義之輪廓表示法 (Semantic-based Contour Representation, SCR) 特徵表達物體、再透過不同的卷積核 (Convolution Kernel) 重釋SC

R特徵並交由全連接類神經網路(Fully Connected Neural Network, FCN) 進行車輛辨識。最後，以多個邊界框 (Bounding Box) 同時檢測前方多台車輛，達到單頁多目標辨識 (Single Shot MultiBox Detector，SSD) 之功能，而邊界框之座標可以透視法 (Perspective View) 計算前車相對距離。根據本研究之實驗結果，在相機以每秒90張影像攝影速度以及影像解析度在640×480像素的條件下，本研究僅須3.61us即可完成單台車輛辨識，車輛辨識率可達到94%，且車輛與非車輛至少保持38%以上之分離度，有效減少感測錯誤的情況

發生。因此，實現一真正高性能、低功耗以及體積小之前方車輛辨識晶片。