1 # Configuration of the image library (included by cf/sherlock)
3 ######## General parameters #####################################################
7 # Default tracing level (0 to disable)
10 # Limits for image allocation
11 ImageMaxDim 0xffff # Maximum width/height (at most 64k-1)
12 ImageMaxBytes 256M # Maximum size in bytes
16 #if CONFIG_IMAGES_DUP || CONFIG_IMAGES_SIM
17 ######## Image signatures #######################################################
21 # To find similar images, Sherlock uses comparison based on regions.
22 # First of all, the imagesim analyser extracts various region features.
23 # Sets of these features are called "image signatures" and they are stored
24 # in the `H' attribute of image objects.
26 # Signatures are later processed by the indexer to build an effective
27 # search structure finally used by the search server. See Indexer and Search
28 # sections for more options.
30 # Minimum image size to apply segmentation. Smaller images are always
31 # compared by the simple "average" method (see ImageSig.CompareMethod).
35 # List of subdivision thresholds in the first phase of segmentation.
36 # Lower the values to increase the average number of regions and vice versa.
37 PreQuantThresholds 6 12 15 20 25 25 30 30 40 40 50 50 60 60 60
39 # Settings for the second phase of segmentation -- usually not so important.
40 # We use an iterative algorithm to improve the average error from the first phase.
41 # We stop the process after PostQuantMaxSteps or if we get only PostQuantThreshold
42 # percentual improvement over the previous step.
47 # BorderBonus and BorderSize parameters can increase or decrease the weight
48 # of image pixels near the borders. Weight of all pixels more than BorderSize * MIN(cols, rows)
49 # pixels far from the edges is 128. Then this value continuously decreses/increses up to 128+BorderSize.
53 # Scaling constants for computation of normalized i-th order inertia features (I1, I2, I3).
54 InertiaScale 2 0.5 0.05
56 # Threshold for detecting textured images (see images/sig-txt.c for details).
57 # Decrease the threshold if you want less detected textures,
58 # set it to zero to disable the algorithm completely.
59 TexturedThreshold 0.32
61 # Signature comparison method:
64 # based on: James Z. Wang, Jia Li and Gio Wiederhold,
65 # "SIMPLIcity: Semantics-Sensitive Integrated Matching for Picture Libraries",
66 # IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23, no. 9, pp. 947-963, 2001.
68 # fuzzy (unstable and unbalanced parameters)
69 # based on: Yixin Chen and James Z. Wang,
70 # "A Region-Based Fuzzy Feature Matching Approach to Content-Based Image Retrieval",
71 # IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 24, no. 9, pp. 1252-1267, 2002.
74 # Simple distance of image features averages (ignores segmentation).
76 CompareMethod integrated
78 # Array of multiplicative constants in feature vector distance computation
79 # (L, u, v, LH, HL, HH, I1, I2, I3, X, Y). Each one must be an integer in range 0..15, default is 4.
80 CompareFeaturesWeights 4 6 6 4 4 4 4 4 4 4 4
85 #ifdef CONFIG_IMAGES_DUP
86 ######## Duplicate finder #######################################################
89 # Detection of image duplicates does not work yet.