Fire Detection Algorithms Using Multimodal ... - Bilkent University
Fire Detection Algorithms Using Multimodal ... - Bilkent University Fire Detection Algorithms Using Multimodal ... - Bilkent University
LIST OF FIGURESxiv2.6 Three-state Markov models for a) flame and b) non-flame movingflame-colored pixels. . . . . . . . . . . . . . . . . . . . . . . . . . 232.7 (a) A child with a fire-colored t-shirt, and b) the absolute sum ofspatial wavelet transform coefficients, |I lh (k, l)|+|I hl (k, l)|+|I hh (k, l)|,of the region bounded by the indicated rectangle. . . . . . . . . . 262.8 (a) Fire, and (b) the absolute sum of spatial wavelet transformcoefficients, |I lh (k, l)|+|I hl (k, l)|+|I hh (k, l)|, of the region boundedby the indicated rectangle. . . . . . . . . . . . . . . . . . . . . . . 262.9 (a) With the method using color and temporal variation only(Method-2.2) [64], false alarms are issued for the fire colored lineon the moving truck and the ground, (b) our method (Method-2.1)does not produce any false alarms. . . . . . . . . . . . . . . . . . 282.10 Sample images (a) and (b) are from Movies 7 and 9, respectively.(c) False alarms are issued for the arm of the man with the methodusing color and temporal variation only (Method-2.2) [64] and (d)on the fire-colored parking car. Our method does not give any falsealarms in such cases (see Table 2.2). . . . . . . . . . . . . . . . . . 292.11 Sample images (a) and (b) are from Movies 2 and 4, respectively.Flames are successfully detected with our method (Method-2.1) in(c) and (d). In (c), although flames are partially occluded by thefence, a fire alarm is issued successfully. Fire pixels are painted inbright green. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.1 Two relatively bright moving objects in FLIR video: a) fire image,and b) a man (pointed with an arrow). Moving objects aredetermined by the hybrid background subtraction algorithm of [19]. 393.2 Equally spaced 64 contour points of the a) walking man, and b)the fire regions shown in Fig. 3.1. . . . . . . . . . . . . . . . . . . 40
LIST OF FIGURESxv3.3 Single-stage wavelet filter bank. The high-pass and the low-passfilter coefficients are {− 1, 1, − 1} and { 1, 1, 1 }, respectively. . . . . 404 2 4 4 2 43.4 The absolute values of a) high-band (wavelet) and b) low-band coefficientsfor the fire region. . . . . . . . . . . . . . . . . . . . . . 413.5 The absolute a) high-band (wavelet) and b) low-band coefficientsfor the walking man. . . . . . . . . . . . . . . . . . . . . . . . . . 423.6 Flicker frequency distributions for a) 10 fps color video, b) 25 fpscolor video and c) 10 fps IR video. These frequency distributionswere obtained by analyzing the temporal variations in the red channelvalue of a pixel at the boundary of a flame region in color-videoclips recorded at 10 fps and 25 fps and intensity value of a pixel atthe boundary of a flame region in an IR video clip recorded at 10fps, respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443.7 Three-state Markov models for a) flame and b) non-flame movingpixels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453.8 Image frames from some of the test clips. a), b) and c) Fire regionsare detected and flame boundaries are marked with arrows. d), e)and f) No false alarms are issued for ordinary moving bright objects. 473.9 Image frames from some of the test clips with fire. Pixels on theflame boundaries are successfully detected. . . . . . . . . . . . . . 484.1 Image frame with smoke and its single level wavelet sub-images.Blurring in the edges is visible. The analysis is carried out in smallblocks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574.2 Single-stage wavelet filter bank. . . . . . . . . . . . . . . . . . . . 594.3 Three-state Markov models for smoke(left) and non-smoke movingpixels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
- Page 3 and 4: I certify that I have read this the
- Page 7 and 8: viiiçin de işaret işleme yöntem
- Page 9: ixThis work was supported in part b
- Page 16 and 17: LIST OF FIGURESxvi4.4 Two moving ob
- Page 18 and 19: List of Tables1.1 Organization of t
- Page 20 and 21: Chapter 1IntroductionDynamic textur
- Page 22: CHAPTER 1. INTRODUCTION 3No fire de
- Page 27 and 28: Chapter 2Flame Detection in Visible
- Page 29 and 30: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 31 and 32: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 33 and 34: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 35 and 36: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 37 and 38: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 39 and 40: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 41 and 42: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 43 and 44: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 45 and 46: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 47 and 48: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 49 and 50: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 51 and 52: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 53 and 54: CHAPTER 2. FLAME DETECTION IN VISIB
- Page 55 and 56: CHAPTER 3. FLAME DETECTION IN INFRA
- Page 57 and 58: CHAPTER 3. FLAME DETECTION IN INFRA
- Page 59 and 60: CHAPTER 3. FLAME DETECTION IN INFRA
- Page 61 and 62: CHAPTER 3. FLAME DETECTION IN INFRA
- Page 63 and 64: CHAPTER 3. FLAME DETECTION IN INFRA
LIST OF FIGURESxv3.3 Single-stage wavelet filter bank. The high-pass and the low-passfilter coefficients are {− 1, 1, − 1} and { 1, 1, 1 }, respectively. . . . . 404 2 4 4 2 43.4 The absolute values of a) high-band (wavelet) and b) low-band coefficientsfor the fire region. . . . . . . . . . . . . . . . . . . . . . 413.5 The absolute a) high-band (wavelet) and b) low-band coefficientsfor the walking man. . . . . . . . . . . . . . . . . . . . . . . . . . 423.6 Flicker frequency distributions for a) 10 fps color video, b) 25 fpscolor video and c) 10 fps IR video. These frequency distributionswere obtained by analyzing the temporal variations in the red channelvalue of a pixel at the boundary of a flame region in color-videoclips recorded at 10 fps and 25 fps and intensity value of a pixel atthe boundary of a flame region in an IR video clip recorded at 10fps, respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443.7 Three-state Markov models for a) flame and b) non-flame movingpixels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453.8 Image frames from some of the test clips. a), b) and c) <strong>Fire</strong> regionsare detected and flame boundaries are marked with arrows. d), e)and f) No false alarms are issued for ordinary moving bright objects. 473.9 Image frames from some of the test clips with fire. Pixels on theflame boundaries are successfully detected. . . . . . . . . . . . . . 484.1 Image frame with smoke and its single level wavelet sub-images.Blurring in the edges is visible. The analysis is carried out in smallblocks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574.2 Single-stage wavelet filter bank. . . . . . . . . . . . . . . . . . . . 594.3 Three-state Markov models for smoke(left) and non-smoke movingpixels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60