- 上一篇：高光谱数据稀疏约束分类识别技术研究
- 下一篇：Python+Django医院病人护理辅助系统设计
摘要行人检测是对象检测中的一个特例。为了检测多个对象，基于霍夫变换的方 法，采用非极大值抑制或模式寻求以霍夫图像定位和区分峰值。这种后处理需要 额外的参数调整，尤其是当应该关注的对象往往是位置相近。在本文中，我们开 发了一个新的概率框架，在许多方面和霍夫变换相关联，分享它的简单性和广泛 的适用性。同时，框架绕过霍夫图像多重峰值识别的问题，并允许多个对象的检 测，而不调用非极大值抑制启发式。因此，实验对于经典的直线检测任务和更先 进的分类水平的行人检测问题，检测准确度上表现出更显著的改进。我们对夜间 行人数据集进行图像预处理，以便使检测结果更为准确,比较预处理方法的优劣。68477
毕业论文关键词 夜间行人检测 霍夫变换 贪心算法 图像预处理
Title The Collection ,Labeling and Test of the Pedestrian Database: the Collection part
Abstract
The pedestrian detection is a special case of object detection. The goal of the pedestrian detection is to get the location of each pedestrian in a frame in the image space. To detect multiple objects of interest, the methods based on Hough transform use non-maxima supression or mode seeking in order to locate and to distinguish peaks in Hough images. Such postprocessing requires tuning of extra parameters and is often fragile, especially when objects of interest tend to be closely located. In the paper, we develop a new probabilistic framework that is in many ways related to Hough transform, sharing its simplicity and wide applicability. At the same time, the framework bypasses the problem of multiple peaks identification in Hough images, and permits detection of multiple objects without invoking non-maximum suppression heuristics. As a result, the experiments demonstrate a significant improvement in detection accuracy both for the classical task of straight line detection and for a more modern category-level (pedestrian) detection problem. We preprocess the images of the pedestrians at night-time. Finally ,we compare the results of each image-preprocess method.
Keywords Pedestrian detection at night-time Hough-Transforms Greedy algorithm Image preprocess
目 次
1 绪论 1
1.1 课题背景及意义 1
1.3 行人检测面临的问题 2
1.4 本文的组织安排 3
2 霍夫变换 4
2.1 霍夫变换基本原理 4
2.2 霍夫森林 5
3 基于概率模型的霍夫变换 6
3.1 霍夫变换的分析 6
3.2 概率构架 7
3.3 用贪心算法求解最大后验概率 8
4 实验成果展示 11
4.1 直线检测 11
4.2 行人检测 13
5 对夜间行人检测的改进 17
5.1 使用 IrfanView 校正色彩 17