Development of Crowd Behavior Analysis Techniques / (Record no. 615911)
[ view plain ]
| 000 -LEADER | |
|---|---|
| fixed length control field | 04415nam a22001697a 4500 |
| 003 - CONTROL NUMBER IDENTIFIER | |
| control field | NUST |
| 082 ## - DEWEY DECIMAL CLASSIFICATION NUMBER | |
| Classification number | 005.1,JAV |
| 100 ## - MAIN ENTRY--PERSONAL NAME | |
| Personal name | Javid, Rakhshanda |
| 9 (RLIN) | 124511 |
| 245 ## - TITLE STATEMENT | |
| Title | Development of Crowd Behavior Analysis Techniques / |
| Statement of responsibility, etc. | Rakhshanda Javid |
| 260 ## - PUBLICATION, DISTRIBUTION, ETC. | |
| Place of publication, distribution, etc. | Rawalpindi, |
| Name of publisher, distributor, etc. | MCS (NUST), |
| Date of publication, distribution, etc. | 2019 |
| 300 ## - PHYSICAL DESCRIPTION | |
| Extent | xiv, 116 p |
| 505 ## - FORMATTED CONTENTS NOTE | |
| Formatted contents note | Automation of behavior analysis of crowds is a challenging task due to the density<br/>variations, occlusions, context, and scene variability. This dissertation aims to address<br/>different challenges in image as well as videos of different crowd systems. The main<br/>goal is to help quantise the crowd behavior, segment salient crowd groups and detect<br/>false crowds in forged images using spatial and temporal cues derived from the crowd<br/>itself.<br/>An improved crowd coherent collectiveness descriptor for analyzing and measuring<br/>the collective motion of crowds is proposed. The descriptor is estimated based on velocities<br/>which can result in zero collectiveness for static crowds. To overcome this<br/>issue a moving weighted average concept is applied for computation of horizontal and<br/>vertical velocities. A voting based scheme is used to cluster the crowd. It takes the information<br/>of clustering from previous frames and based on the maximum voting results<br/>cluster the crowd at current time. The proposed scheme is useful for density estimation<br/>and behavior analysis of different crowds and its application for creatures (sea-birds,<br/>fish, and others etc.) is also explored. Visual and quantitative analysis verifies the<br/>significance of the proposed scheme.<br/>To quantify the crowd behavior we use different descriptors. The existing descriptors<br/>are contextual and generally provide information about crowd density categorized as<br/>low, medium or high. However, other properties of the crowd like speed, direction,<br/>shape and merging probabilities (of different crowds at group level) are also important<br/>for crowd analysis. In the proposed technique, crowd descriptors are introduced which<br/>are robust against different outliers and different densities of the crowd. Simulations<br/>on various datasets show the applicability of proposed descriptors.<br/>Crowd flow and dominant motion detection technique (locally between the clusters)<br/>are combined to detect crowd saliency in this research. For each cluster, the<br/>saliency map is computed using motion and contrast cues. The dominant motion is<br/>detected using k means clustering. The proposed technique is able to segment the<br/>crowd cluster with maximum motion and can even detect the straight-line motions.<br/>The proposed scheme can identify salient crowd groups. The applicability of the proposed<br/>scheme has been demonstrated for microscopic medical data. The motion of<br/>the immune system makes it more salient than the surrounding cells. Thus exploiting<br/>this idea a spatio-temporal technique is proposed. In which temporal saliency is computed<br/>using extended Lucas Kanade and coherent clustering while spatial saliency is<br/>computed on feature extraction. Simulations on different datasets show the effectiveness/<br/>applicability of the proposed technique.<br/>The authenticity and reliability of digital images has become one of the major coni<br/>cerns recently due to the ease in manipulating and modifying these images. Similar<br/>manipulation in crowded images give rise to false crowd, where a person or group of<br/>persons is copied and pasted in the same image. Thus, the detection of such false<br/>crowds is the focus of current research. In this research, false crowd detection in forged<br/>images is carried out using a modified and improved PatchMatch algorithm which can<br/>even detect multiple copies of the same instance. To separate humans from non-human<br/>objects a human detection algorithm is used in the post-processing phase. A benchmark<br/>database consisting of false crowd images has also been developed. Experimental results<br/>confirm that the technique is capable of detecting the false crowds successfully<br/>and is even robust for multiple cloning problem.<br/>We experimentally demonstrate the effectiveness and robustness of proposed algorithms<br/>by quantifying crowds at the group level, segment salient crowd groups and<br/>detecting false crowd groups. |
| 650 ## - SUBJECT ADDED ENTRY--TOPICAL TERM | |
| Topical term or geographic name entry element | PhD Computer Software Engineering Thesis |
| 9 (RLIN) | 132801 |
| 651 ## - SUBJECT ADDED ENTRY--GEOGRAPHIC NAME | |
| Geographic name | PhD CSE Thesis |
| 9 (RLIN) | 132802 |
| 700 ## - ADDED ENTRY--PERSONAL NAME | |
| Personal name | Supervised by Dr. Naveed Iqbal Rao |
| 9 (RLIN) | 132893 |
| 942 ## - ADDED ENTRY ELEMENTS (KOHA) | |
| Source of classification or shelving scheme | |
| Koha item type | Thesis |
| Withdrawn status | Lost status | Source of classification or shelving scheme | Damaged status | Not for loan | Permanent Location | Current Location | Shelving location | Date acquired | Total Checkouts | Full call number | Barcode | Date last seen | Price effective from | Koha item type | Public note |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Military College of Signals (MCS) | Military College of Signals (MCS) | Thesis | 01/26/2026 | 005.1,JAV | MCSPhD CS-09 | 01/26/2026 | 01/26/2026 | Thesis | Almirah No.68, Shelf No.5 |