《序列圖像中的目標分析技術》是李子印,孫志海編輯出版的圖書。
基本信息,內容簡介,目錄信息,
基本信息
序列圖像中的目標分析技術
作 譯 者：李子印,孫志海
出版時間：2016-08
千 字 數：313
版 次：01-01
頁 數：248
開 本：16開
I S B N ：9787121297632
內容簡介
本書以序列圖像中目標分析技術的基本過程為主線，系統地介紹了目標分析的基本理論，詳細講解了作者的研究成果。緒論重點對序列圖像中目標分析技術的研究現狀進行了分析。目檢測部分，提出了一種基本的視頻運動目標檢測技術框架；在此基礎上提出了兩種改進的目標檢測算法，可分別用於需要精確檢測目標和閾值化後目標連通性較差的套用場合；針對幀間差分法的不足，提出了一種基於差分背景融合建模的目標檢測算法。目標定位部分，提出了一種基於減法聚類算法的目標定位技術和一種橢圓域減法聚類目標定位方法；提出了減法聚類目標定位算法的七點最佳化技術；另外，提出了一種基於非參數核密度估計的目標定位技術，可根據套用靈活選擇核函式估計樣本點的密度值分布；針對減法聚類技術複雜度高的問題，提出了一種基於Nystr?m密度值逼近的減法聚類方法。目標運動估計部分，為了降低運動估計的計算複雜度，提出了一種基於運動場預測的六邊形塊運動估計算法和一種基於運動場預測的部分失真運動估計算法；另外，對UMHexagonS算法進行了改進。目標跟蹤與識別部分，針對複雜背景下的目標跟蹤，提出了一種基於圖像感知哈希技術的目標跟蹤算法；針對遮擋情況，提出了一種自適應步長選擇的NCC圖像匹配算法；最後，採用基於團塊和軌跡分析的方法實現了區域入侵、人體跌倒、遺留物檢測、人體徘徊四種異常行為的判定。
目錄信息
第1 章 緒論··········································································································· 1
1.1 研究背景及意義······························································································· 1
1.2 視頻運動目標檢測研究現狀··········································································· 3
1.2.1 背景差法··································································································· 4
1.2.2 鄰幀差法··································································································· 5
1.2.3 光流法······································································································· 5
1.3 視頻運動目標定位研究現狀··········································································· 6
1.4 視頻運動估計研究現狀··················································································· 7
1.5 視頻運動目標跟蹤研究現狀··········································································· 8
1.6 本書的內容及章節安排················································································· 10
1.6.1 本書的內容····························································································· 10
1.6.2 本書的章節安排······················································································11
1.7 本章小結········································································································ 13
參考文獻················································································································ 13
第2 章 基於積累差異背景建模的視頻運動目標檢測······························21
2.1 引言················································································································ 21
2.2 基於積累差異的背景建模············································································· 23
2.2.1 積累差異································································································· 23
2.2.2 積累差異背景建模················································································· 23
2.3 Otsu 自適應閾值化及目標輪廓提取····························································· 25
2.3.1 Otsu 閾值化算法···················································································· 25
2.3.2 改進的Otsu 閾值化算法········································································ 26
2.3.3 目標輪廓提取························································································· 27
2.4 兩步區域生長目標連通區域標記································································· 27
2.5 目標質心關聯································································································ 28
2.5.1 質心標記································································································· 28
2.5.2 質心關聯································································································· 28
2.6 監控場合行人及運動車輛檢測實驗····························································· 29
2.6.1 積累差異背景建模及運動目標檢測······················································ 29
2.6.2 運動目標輪廓提取及質心關聯····························································· 31
2.7 夜間運動車輛檢測實驗················································································· 32
2.8 語義視頻運動目標檢測實驗········································································· 36
2.8.1 顏色空間及膚色模型············································································· 36
2.8.2 實驗效果及分析····················································································· 38
2.9 積累差異背景建模與GMM 背景建模的比較實驗······································ 40
2.10 本章小結······································································································ 43
參考文獻················································································································ 43
第3 章 基於差分背景融合建模的運動目標檢測·······································46
3.1 引言················································································································ 46
3.2 算法基本思想································································································ 46
3.3 背景模型的建立····························································································· 47
3.4 自適應背景更新策略····················································································· 48
3.5 實驗效果與分析····························································································· 50
3.6 本章小結········································································································ 54
參考文獻················································································································ 55
第4 章 融合Knockout 摳圖技術的視頻運動目標檢測····························56
4.1 引言················································································································ 56
4.2 摳圖技術簡介······································································································56
4.3 視頻運動目標區域初步檢測········································································· 58
4.4 Knockout 視頻運動目標檢測········································································ 58
4.4.1 Knockout 技術························································································ 58
4.4.2 透明度計算····························································································· 59
4.4.3 自動區域標記························································································· 60
4.5 實驗效果及分析····························································································· 61
4.6 本章小結········································································································ 62
參考文獻················································································································ 63
第5 章 基於格線區域劃分的視頻運動目標檢測·······································64
5.1 引言················································································································ 64
5.2 算法設計········································································································ 64
5.3 積累差異背景建模及運動檢測·········································································65
5.4 格線區域劃分································································································ 65
5.5 實驗測試及分析····························································································· 67
5.5.1 空域連通性差時的檢測情況································································· 68
5.5.2 格線區域劃分前後檢測效果比較························································· 69
5.5.3 與GMM 法檢測效果的比較································································· 71
5.5.4 閾值Tg 對檢測效果的影響···································································· 72
5.5.5 格線大小對檢測效果影響的討論························································· 73
5.6 不同檢測方法的分析及比較········································································· 75
5.6.1 檢測效果對比························································································· 75
5.6.2 計算複雜度分析····················································································· 77
5.6.3 算法實用性說明····················································································· 77
5.7 本章小結··············································································································78
參考文獻·······················································································································78
第6 章 基於減法聚類算法的視頻運動目標定位·······································79
6.1 引言················································································································ 79
6.2 山峰聚類算法································································································ 80
6.3 減法聚類算法······································································································82
6.4 減法聚類目標定位······························································································83
6.5 減法聚類目標定位實驗················································································· 84
6.5.1 減法聚類目標定位過程········································································· 84
6.5.2 與區域生長目標定位的比較································································· 85
6.5.3 減法聚類定位的抗噪性實驗································································· 87
6.6 橢圓域減法聚類目標定位············································································· 88
6.6.1 橢圓域減法聚類算法············································································· 88
6.6.2 橢圓域減法聚類目標定位實驗····························································· 89
6.7 橢圓域減法聚類定位套用舉例····································································· 93
6.8 本章小結········································································································ 95
參考文獻················································································································ 96
第7 章 視頻目標定位的減法聚類改進算法················································97
7.1 引言················································································································ 97
7.2 減法聚類目標定位························································································· 98
7.3 算法最佳化········································································································ 99
7.3.1 考慮不同維度的鄰域半徑··································································· 100
7.3.2 引入下採樣技術··················································································· 100
7.3.3 選擇合適的密度值函式······································································· 101
7.3.4 構造格線重定義數據集······································································· 103
7.3.5 模糊隸屬度前景像素聚類··································································· 105
7.3.6 修正目標鄰域半逕取值······································································· 105
7.3.7 確定視頻運動目標尺度和方向··························································· 106
7.4 實驗效果及分析··························································································· 107
7.4.1 減法聚類算法視頻目標定位過程······················································· 108
7.4.2 有效鄰域半徑對定位結果的影響······················································· 109
7.4.3 下採樣定位效果··················································································· 109
7.4.4 不同密度函式的定位情況····································································110
7.4.5 模糊隸屬度前景像素歸類實驗····························································111
7.4.6 目標尺度和方向參數確定實驗····························································111
7.5 本章小結·······································································································115
參考文獻···············································································································115
第8 章 非參數核密度估計視頻目標空域定位·········································117
8.1 引言···············································································································117
8.2 非參數核密度估計························································································118
8.3 非參數核密度估計視頻目標空域定位························································119
8.4 實驗結果與分析··························································································· 120
8.4.1 定位過程分析······················································································· 120
8.4.2 不同核函式定位效果分析··································································· 122
8.4.3 不同定位算法效果比較······································································· 123
8.4.4 頻寬對定位的影響··············································································· 124
8.5 結論·············································································································· 125
參考文獻·············································································································· 126
第9 章 基於Nystr?m 密度值逼近的減法聚類········································127
9.1 引言·············································································································· 127
9.2 考慮不同維度鄰域的減法聚類··································································· 129
9.3 Nystr?m 樣本密度值逼近··········································································· 129
9.3.1 Nystr?m 樣本相關係數加權······························································· 130
9.3.2 減法聚類樣本密度值逼近··································································· 131
9.3.3 Nystr?m 減法聚類小結······································································· 132
9.4 實驗結果和分析··························································································· 133
9.4.1 人工數據實驗······················································································· 133
9.4.2 彩色圖像聚類實驗··············································································· 138
9.4.3 UCI 數據集聚類實驗··········································································· 140
9.5 結論·············································································································· 141
參考文獻·············································································································· 141
第10 章 基於運動場預測的六邊形塊運動估計·······································143
10.1 引言············································································································ 143
10.2 經典快速塊運動估計搜尋算法································································· 144
10.2.1 MPEG—4 校驗模型採用的菱形搜尋算法（DS）··························· 144
10.2.2 PMVFAST 算法·················································································· 144
10.3 PMVHEXBS 算法······················································································ 146
10.3.1 運動估計算法中的搜尋模式····························································· 146
10.3.2 本章算法搜尋模式的選擇································································· 147
10.3.3 視頻序列中運動的時空相關性························································· 149
10.3.4 初始搜尋點的確定············································································· 150
10.3.5 “足夠好就停止搜尋”思想的套用······················································ 151
10.3.6 改進的部分失真準則（MPDC） ···················································· 152
10.3.7 PMVHEXBS 算法的步驟··································································· 153
10.3.8 本章算法分析····················································································· 154
10.4 實驗結果···································································································· 156
10.5 本章小結 ··································································································· 160
參考文獻·············································································································· 161
第11 章 基於運動信息自適應的快速運動估計·······································163
11.1 引言············································································································ 163
11.2 UMHexagonS 算法介紹············································································· 164
11.3 運動信息自適應的快速運動估計算法····················································· 165
11.3.1 塊尺寸自適應的動態搜尋窗的計算·················································· 165
11.3.2 運動類型自適應的搜尋方案選擇······················································ 166
11.3.3 帶方向的十字型搜尋········································································· 166
11.3.4 自適應的矩形-菱形搜尋···································································· 167
11.3.5 運動方向自適應的多層次八邊形區域搜尋······································ 168
11.3.6 塊尺寸自適應的六邊形搜尋····························································· 168
11.4 實驗仿真與分析························································································· 169
11.5 結束語········································································································ 171
參考文獻·············································································································· 172
