sqed 0.4.1 → 0.4.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: d47942c64d4e9dfee4eefdb864de8b537aad0166
-  data.tar.gz: e6a282b0b85f83615315c291bf2cbf66fe26b15c
+  metadata.gz: 047c0529db7c52677c9ece993cbfbb76aad0e475
+  data.tar.gz: d700a6fffb703f94bcb1b5e0c3250b16b9b65894
 SHA512:
-  metadata.gz: e53e4bbc42260cd2723eaa4ef917a2f8b1c99b8947bb3af3675beeffcee07ad6567c753a601a38b82e0ae86d27e103bd6b06de06d8517d38394b228dc606c64b
-  data.tar.gz: 861e69ad4e87e8e7261f3020f4d9aef5c1ce1a2bf3962dae1ac4eef898ac5ddf27f85d3a53c3a448e04bd401305be79c5a014a0f5f9b099ea9f8fb94b1f23261
+  metadata.gz: c45f312e2c5d42706104997a101b9aa5bcc25b37ddd774260b337e16d5426f17bf75614fd17b662c8444beecdd5c79e9d7520060b98a665f90a61f830aeec599
+  data.tar.gz: 0c14d57039134180aa3f9f85d80e24fe3d4863eb4a1e6116e2f6bfd08d4e18ad8d031f7de7d0cf3cd87625db2f4f7bb27edfc034cee1b1c0962c3ca64a67cf0b

data/lib/sqed.rb

@@ -4,6 +4,7 @@ recent_ruby = RUBY_VERSION >= '2.4.1'
 raise 'IMPORTANT: sqed gem requires ruby >= 2.4.1' unless recent_ruby
 
 require 'rmagick'
+require 'sqed_utils'
 
 # Instances take the following
 # 1) An :image @image

data/lib/sqed/boundary_finder.rb

@@ -119,8 +119,9 @@ class Sqed
     # @param scan
     #   (:rows|:columns), :rows finds vertical borders, :columns finds horizontal borders
     #
+    #
+    # image: image, sample_subdivision_size: nil, sample_cutoff_factor: nil, scan: :rows, boundary_color: :green)
     def self.color_boundary_finder(**opts)
-      # image: image, sample_subdivision_size: nil, sample_cutoff_factor: nil, scan: :rows, boundary_color: :green)
       image = opts[:image]
       sample_subdivision_size = opts[:sample_subdivision_size]
       sample_cutoff_factor = opts[:sample_cutoff_factor]
@@ -222,54 +223,6 @@ class Sqed
       [hit_ranges.first, hit_ranges[(hit_ranges.length / 2).to_i], hit_ranges.last]
     end
 
-    # @return [Array]
-    #    like `[0,1,2]`
-    # If median-min or max-median * width_factor are different from one another (by more than width_factor) then replace the larger wth the median +/- 1/2 the smaller
-    # Given [10, 12, 20] and width_factor 2 the result will be [10, 12, 13]
-    #
-    def corrected_frequency(frequency_stats, width_factor = 3.0, max_width = nil)
-      v0 = frequency_stats[0]
-      m = frequency_stats[1]
-      v2 = frequency_stats[2]
-
-      width_pct = nil
-
-      # If the width of the detected line is > 5 percent then
-      # assume the center is good measure, and use an arbitrary percentage width.
-      if !max_width.nil?
-        width_pct = (v2 - v0).to_f / max_width.to_f
-      end
-
-      if !width_pct.nil? && width_pct >= 0.05
-        half_width = max_width * 0.025
-        z = [m - half_width, m, m + half_width]
-        return z
-      end
-
-      #  a = m - v0
-      #  b = v2 - m
-
-      #  largest = (a > b ? a : b)
-
-      #  c = a * width_factor
-      #  d = b * width_factor
-
-      #  if c > largest || d > largest
-      #    e = c > largest ? ((m - b) / 2).to_i : v0
-      #    f = d > largest ? ((m + a) / 2).to_i : v2
-      #    [e, m, f]
-      #  else
-      #    frequency_stats
-      #  end
-
-      if a * width_factor > largest
-        [(m - (v2 - m) / 2).to_i, m, v2]
-      elsif b * width_factor > largest
-        [v0, m, (m + (m - v0) / 2).to_i]
-      else
-        frequency_stats
-      end
-    end
 
     # Returns an Integer, the maximum of the pairwise differences of the values in the array
     # For example, given

data/lib/sqed/boundary_finder/color_line_finder.rb

@@ -30,7 +30,57 @@ class Sqed::BoundaryFinder::ColorLineFinder < Sqed::BoundaryFinder
   def find_bands
     case layout    # boundaries.coordinates are referenced from stage image
 
-      # No specs for this yet
+    # No specs for this yet
+    when :lep_stage
+      top_bottom_split = Sqed::BoundaryFinder.color_boundary_finder(image: image, scan: :columns, boundary_color: boundary_color)              # detect vertical division [array]
+      left_right_split = Sqed::BoundaryFinder.color_boundary_finder(image: image, sample_subdivision_size: 2, boundary_color: boundary_color)  # detect horizontal division [array]
+
+      boundaries.set(6, [0, top_bottom_split[2], left_right_split[0], image.rows - top_bottom_split[2]] )
+
+      left_top_image = image.crop( 0, 0, left_right_split[0], top_bottom_split[0], true) 
+
+      left_top_split = 
+        SqedUtils.corrected_frequency(
+          Sqed::BoundaryFinder.color_boundary_finder(image: left_top_image, boundary_color: boundary_color),
+          max_width: left_top_image.columns,
+          width_factor: 1.8
+      )
+
+      boundaries.set(0, [0, 0, left_top_split[1], top_bottom_split[0]] ) # keep as 1 for safety
+     
+      boundaries.set(1, [left_top_split[2], 0, left_top_image.columns - left_top_split[2], top_bottom_split[0]] ) # <
+
+      boundaries.set(2, [left_right_split[2], 0, image.columns - left_right_split[0], top_bottom_split[0]] )
+
+      bottom_right_image =  image.crop(left_right_split[2], top_bottom_split[2], image.columns - left_right_split[2], image.rows - top_bottom_split[2], true)
+
+      bottom_right_split =
+        SqedUtils.corrected_frequency(
+          Sqed::BoundaryFinder.color_boundary_finder(image: bottom_right_image, boundary_color: boundary_color, scan: :rows),
+          max_width: bottom_right_image.columns,
+          width_factor: 1.8
+      )
+
+      boundaries.set(3, [left_right_split[2] + bottom_right_split[2], top_bottom_split[2], image.columns - bottom_right_split[2], image.rows - top_bottom_split[2] ] )
+
+      bottom_right_left_image = image.crop(left_right_split[2], top_bottom_split[2], bottom_right_split[0], image.rows - top_bottom_split[2], true)  
+
+      bottom_right_left_split =
+        SqedUtils.corrected_frequency(
+          Sqed::BoundaryFinder.color_boundary_finder(image: bottom_right_left_image, scan: :columns, boundary_color: boundary_color),
+          max_width: bottom_right_left_image.columns,
+          width_factor: 1.8
+      )
+
+      boundaries.set(4, [left_right_split[2], top_bottom_split[2], bottom_right_image.columns - bottom_right_left_image.columns, bottom_right_left_split[0] ] )
+
+      boundaries.set(5, [
+        left_right_split[2], 
+        top_bottom_split[2] + bottom_right_left_split[2],
+        bottom_right_image.columns - bottom_right_left_image.columns,
+        bottom_right_left_image.rows - bottom_right_left_split[2]  
+      ])
+
     when :seven_slot
       top_bottom_split = Sqed::BoundaryFinder.color_boundary_finder(image: image, scan: :columns, boundary_color: boundary_color)              # detect vertical division [array]
       left_right_split = Sqed::BoundaryFinder.color_boundary_finder(image: image, sample_subdivision_size: 2, boundary_color: boundary_color)  # detect horizontal division [array]
@@ -41,12 +91,19 @@ class Sqed::BoundaryFinder::ColorLineFinder < Sqed::BoundaryFinder
       right_top_image = image.crop( left_right_split[2], 0, image.columns - left_right_split[2], top_bottom_split[0] , true) # sections 1,2
       right_bottom_image = image.crop(left_right_split[2], top_bottom_split[2], image.columns - left_right_split[2], image.rows - top_bottom_split[2], true)  # sections 3,4,5
 
-      right_top_split = corrected_frequency(Sqed::BoundaryFinder.color_boundary_finder(image: right_top_image, boundary_color: boundary_color), 1.8, right_top_image.columns) # vertical line b/w 1 & 2, use "corrected_frequency" to account for color bleed from previous crop
+      right_top_split = ::SqedUtils.corrected_frequency(
+        Sqed::BoundaryFinder.color_boundary_finder(image: right_top_image, boundary_color: boundary_color),
+        width_factor: 1.8,
+        max_width: right_top_image.columns
+      ) # vertical line b/w 1 & 2, use "corrected_frequency" to account for color bleed from previous crop
 
       boundaries.set(1, [left_right_split[2], 0, right_top_split[0], top_bottom_split[0]] )
       boundaries.set(2, [left_right_split[2] + right_top_split[2], 0, right_top_image.columns - right_top_split[2], top_bottom_split[0]])
 
-      right_bottom_split = corrected_frequency(Sqed::BoundaryFinder.color_boundary_finder(image: right_bottom_image, scan: :columns, sample_subdivision_size: 2, boundary_color: boundary_color), 1.8, right_bottom_image.rows) # horizontal line b/w (5,3) & 4, use "corrected_frequency" to account for color bleed from previous crop
+      right_bottom_split = SqedUtils.corrected_frequency(
+        Sqed::BoundaryFinder.color_boundary_finder(image: right_bottom_image, scan: :columns, sample_subdivision_size: 2, boundary_color: boundary_color),
+        width_factor: 1.8,
+        max_width: right_bottom_image.rows) # horizontal line b/w (5,3) & 4, use "corrected_frequency" to account for color bleed from previous crop
 
       bottom_right_top_image = right_bottom_image.crop(0,0, image.columns - left_right_split[2], right_bottom_split[0], true) # 3,5
 

data/lib/sqed/version.rb

@@ -1,3 +1,3 @@
 class Sqed
-  VERSION = '0.4.1'.freeze
+  VERSION = '0.4.2'.freeze
 end

data/lib/sqed_config.rb

@@ -1,5 +1,7 @@
 # require 'zxing'
 
+require_relative 'sqed_utils'
+
 require_relative 'sqed/parser'
 require_relative 'sqed/parser/ocr_parser'
 require_relative 'sqed/parser/barcode_parser'
@@ -62,6 +64,11 @@ module SqedConfig
   #    6 |--------
   #      |   4
   #
+  #    0 | 1 | 2
+  #   ------------
+  #      | 5 |      :lep_stage
+  #    6 |---- 3  
+  #      | 4 |
   #
   # Hash values are used to stub out
   # the Sqed::Boundaries instance.
@@ -74,7 +81,8 @@ module SqedConfig
     right_t: [0, 1, 2],
     left_t: [0, 1, 2],
     internal_box: [0],
-    seven_slot: [0, 1, 2, 3, 4, 5, 6]
+    seven_slot: [0, 1, 2, 3, 4, 5, 6],
+    lep_stage: [0, 1, 2, 3, 4, 5, 6]
   }.freeze
 
   # Each element of the layout is a "section".
@@ -142,6 +150,12 @@ module SqedConfig
       boundary_finder: Sqed::BoundaryFinder::ColorLineFinder,
       layout: :seven_slot,
       metadata_map: { 0 => :collecting_event_labels, 1 => :determination_labels, 2 => :other_labels, 3 => :image_registration, 4 => :curator_metadata, 5 => :identifier, 6 => :specimen }
+    },
+
+    lep_stage: {
+      boundary_finder: Sqed::BoundaryFinder::ColorLineFinder,
+      layout: :lep_stage,
+      metadata_map: { 0 => :curator_metadata, 1 => :collecting_event_labels, 2 => :image_registration, 3 => :identifier, 4 => :other_labels, 5 => :determination_labels, 6 => :specimen }
     }
   }.freeze
 

data/lib/sqed_utils.rb

@@ -0,0 +1,60 @@
+require 'byebug'
+
+# Functions that don't belong in Sqed proper
+#
+module SqedUtils
+
+
+  # @return [Array]
+  #    like `[0,1,2]`
+  #
+  # @param frequency_stats [Array]
+  #   like [1,2,3]
+  #
+  # @param max_width [Integer]
+  #   required, the width of the image in question
+  #
+  # See tests. This code does a rough job of smoothing out boundaries that seem to 
+  # be biased on one side or the other.  Definitely could be refined to use a more
+  # weighted approach. 
+  #
+  def self.corrected_frequency(frequency_stats, width_factor: 3.0, max_width: nil)
+
+    return frequency_stats if max_width.nil?
+
+    v0 = frequency_stats[0]
+    m = frequency_stats[1]
+    v2 = frequency_stats[2]
+
+    width_pct = (v2.to_f - v0.to_f) / max_width.to_f
+
+    return frequency_stats if (width_pct * 100) <= 2.0
+
+    a = (m - v0).abs
+    b = (v2 - m).abs
+
+    c = a * width_factor
+    d = b * width_factor
+
+    largest = (a > b ? a : b)
+
+    l = (m - b / 2) 
+    l = 0 if l < 0
+
+    r = (m + a / 2)
+    r = max_width if r > max_width
+
+    x = a * width_factor
+    y = b * width_factor
+
+    [
+      x > largest ? l : v0,
+      m,
+      y > largest ? r : v2
+    ]
+
+  end
+
+
+
+end

data/spec/lib/sqed_utils_spec.rb

@@ -0,0 +1,42 @@
+require 'sqed_utils'
+
+describe SqedUtils do
+
+  context '.corrected_frequency' do
+
+    specify '[1,2,3]' do
+      expect(SqedUtils.corrected_frequency([1,2,3])).to eq([1,2,3] )
+    end
+
+    specify '[1,2,3], max_width: 100' do
+      expect(SqedUtils.corrected_frequency([1,2,3], max_width: 100)).to eq([1,2,3] )
+    end
+
+    specify '[90,100,1000], max_width: 2000' do
+      expect(SqedUtils.corrected_frequency([90,100,1000], max_width: 2000)).to eq([90, 100, 105] )
+    end
+
+    specify '[90,100,1000], max_width: 2000)' do
+      expect(SqedUtils.corrected_frequency([90,100,1000], max_width: 2000)).to eq([90, 100, 105] )
+    end
+
+    specify '[90,1000,1010], max_width: 2000)' do
+      expect(SqedUtils.corrected_frequency([90,1000,1010], max_width: 2000)).to eq([995, 1000, 1010] )
+    end
+
+    specify '[10, 12, 20], max_width: 2000' do
+      expect(SqedUtils.corrected_frequency([10, 12, 20], max_width: 2000)).to eq([10, 12, 20] )
+    end
+
+   specify '[10, 12, 20], max_width: 100)' do
+      expect(SqedUtils.corrected_frequency([10, 12, 20], max_width: 100)).to eq([10, 12, 13] )
+    end
+
+   specify '[10, 12, 20], max_width: 100, width_factor: 5)' do
+     expect(SqedUtils.corrected_frequency([1, 3, 10], max_width: 100, width_factor: 5)).to eq([0, 3, 4] )
+   end
+
+
+  end
+
+end

data/spec/lib/stage_handling/lep_stage_spec.rb

@@ -0,0 +1,64 @@
+require 'spec_helper'
+
+describe 'handling lep stage images' do
+
+  let(:image) { ImageHelpers.lep_stage }
+  let(:sqed) do
+    Sqed.new(
+      image: image,
+      pattern: :lep_stage,
+      boundary_color: :red,
+      has_border: false )
+  end
+
+  let(:m) do
+    { 0 => :curator_metadata,
+      1 => :collecting_event_labels,
+      2 => :image_registration,
+      3 => :identifier,
+      4 => :other_labels,
+      5 => :determination_labels,
+      6 => :specimen 
+    }
+  end
+
+  context 'parses' do
+    specify 'new() without errors' do
+      expect(sqed).to be_truthy
+    end
+
+    specify 'get result without errors' do
+      expect(sqed.result).to be_truthy
+    end
+  end
+
+  context 'trickier boundaries - without thumbnail' do
+    let(:s) { Sqed.new(image: ImageHelpers.lep_stage, use_thumbnail: false, pattern: :lep_stage, boundary_color: :red, has_border: false ) }
+
+    specify 'boundaries are reasonable' do
+      s.result
+      c = s.boundaries.coordinates
+      c.each do |section, values|
+        c[section].each_with_index do |v, i|
+          msg = "section #{section}, index #{i} has a bad value '#{v}'"
+          expect(v > -1).to be_truthy, msg
+        end
+      end
+    end
+  end
+
+  context 'trickier boundaries - with_thumbnail' do
+    let(:s) { Sqed.new(image: ImageHelpers.lep_stage, use_thumbnail: true, pattern: :lep_stage, boundary_color: :red, has_border: false ) }
+
+    specify 'boundaries are reasonable' do
+      s.result
+      c = s.boundaries.coordinates
+      c.each do |section, values|
+        c[section].each_with_index do |v, i|
+          msg = "section #{section}, index #{i} has a bad value '#{v}'"
+          expect(v > -1).to be_truthy, msg
+        end
+      end
+    end
+  end 
+end

data/spec/support/image_helpers.rb

@@ -35,6 +35,10 @@ module ImageHelpers
       get_image 'stage_images/inhs_7_slot.jpg'
     end
 
+    def lep_stage
+      get_image 'stage_images/lep_stage.jpg'
+    end
+
     def inhs_stage_7_slot2
       get_image 'stage_images/inhs_7_slot2.jpg'
     end

data/sqed.gemspec

@@ -8,7 +8,7 @@ Gem::Specification.new do |spec|
   spec.version     = Sqed::VERSION
   spec.authors     = ["Matt Yoder", "Rich Flood"]
   spec.email       = ["diapriid@gmail.com"]
-  spec.summary     = %q{Specimens Quickly Extracted and Digitized, or just "squid". A ruby gem for image related specimen accessioning.}
+  spec.summary     = %q{Specimens Quickly extracted and Digitized, or just "squid". A ruby gem for image based specimen accessioning.}
   spec.description = %q{A utility gem to aid in the processing of images taken in the process of digitizing natural history collections.}
   spec.homepage    = "http://github.com/SpeciesFileGroup/sqed"
   spec.license     = "NCSA"
@@ -18,17 +18,17 @@ Gem::Specification.new do |spec|
   spec.test_files    = spec.files.grep(%r{^(test|spec|features)/})
   spec.require_paths = ["lib"]
 
-  spec.add_dependency 'rake', '~> 12.0'
+  spec.add_dependency 'rake', '~> 12.3'
   spec.add_dependency 'rmagick', '~> 2.16'  
-  spec.add_dependency 'rtesseract', '~> 2.1.0'
+  spec.add_dependency 'rtesseract', '~> 2.2.0'
 
   # A qrcode reader, too many problems with compiling, dependencies
   # spec.add_dependency 'zxing_cpp', '~> 0.1.0'
 
-  spec.add_development_dependency 'rspec', '~> 3.6'
+  spec.add_development_dependency 'rspec', '~> 3.8'
   spec.add_development_dependency 'bundler', '~> 1.5'
   # spec.add_development_dependency 'did_you_mean', '~> 0.9'
-  spec.add_development_dependency 'byebug', '~> 9.0.6'
+  spec.add_development_dependency 'byebug', '~> 10.0.0'
   spec.add_development_dependency 'awesome_print', '~> 1.8'
 end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: sqed
 version: !ruby/object:Gem::Version
-  version: 0.4.1
+  version: 0.4.2
 platform: ruby
 authors:
 - Matt Yoder
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-03-23 00:00:00.000000000 Z
+date: 2018-09-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
@@ -17,14 +17,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '12.0'
+        version: '12.3'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '12.0'
+        version: '12.3'
 - !ruby/object:Gem::Dependency
   name: rmagick
   requirement: !ruby/object:Gem::Requirement
@@ -45,28 +45,28 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.1.0
+        version: 2.2.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.1.0
+        version: 2.2.0
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '3.6'
+        version: '3.8'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '3.6'
+        version: '3.8'
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
@@ -87,14 +87,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 9.0.6
+        version: 10.0.0
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 9.0.6
+        version: 10.0.0
 - !ruby/object:Gem::Dependency
   name: awesome_print
   requirement: !ruby/object:Gem::Requirement
@@ -139,6 +139,7 @@ files:
 - lib/sqed/result.rb
 - lib/sqed/version.rb
 - lib/sqed_config.rb
+- lib/sqed_utils.rb
 - spec/lib/sqed/boundaries_spec.rb
 - spec/lib/sqed/boundary_finder/color_line_finder_spec.rb
 - spec/lib/sqed/boundary_finder/cross_finder_spec.rb
@@ -150,6 +151,8 @@ files:
 - spec/lib/sqed/parser_spec.rb
 - spec/lib/sqed/result_spec.rb
 - spec/lib/sqed_spec.rb
+- spec/lib/sqed_utils_spec.rb
+- spec/lib/stage_handling/lep_stage_spec.rb
 - spec/lib/stage_handling/seven_slot_spec.rb
 - spec/spec_helper.rb
 - spec/support/files/barcode_images/code_128_barcode.png
@@ -175,6 +178,7 @@ files:
 - spec/support/files/stage_images/inhs_7_slot.jpg
 - spec/support/files/stage_images/inhs_7_slot2.jpg
 - spec/support/files/stage_images/inhs_four_thirds.jpg
+- spec/support/files/stage_images/lep_stage.jpg
 - spec/support/files/test0.jpg
 - spec/support/files/test1.jpg
 - spec/support/files/test2.jpg
@@ -206,8 +210,8 @@ rubyforge_project:
 rubygems_version: 2.6.14
 signing_key: 
 specification_version: 4
-summary: Specimens Quickly Extracted and Digitized, or just "squid". A ruby gem for
-  image related specimen accessioning.
+summary: Specimens Quickly extracted and Digitized, or just "squid". A ruby gem for
+  image based specimen accessioning.
 test_files:
 - spec/lib/sqed/boundaries_spec.rb
 - spec/lib/sqed/boundary_finder/color_line_finder_spec.rb
@@ -220,6 +224,8 @@ test_files:
 - spec/lib/sqed/parser_spec.rb
 - spec/lib/sqed/result_spec.rb
 - spec/lib/sqed_spec.rb
+- spec/lib/sqed_utils_spec.rb
+- spec/lib/stage_handling/lep_stage_spec.rb
 - spec/lib/stage_handling/seven_slot_spec.rb
 - spec/spec_helper.rb
 - spec/support/files/barcode_images/code_128_barcode.png
@@ -245,6 +251,7 @@ test_files:
 - spec/support/files/stage_images/inhs_7_slot.jpg
 - spec/support/files/stage_images/inhs_7_slot2.jpg
 - spec/support/files/stage_images/inhs_four_thirds.jpg
+- spec/support/files/stage_images/lep_stage.jpg
 - spec/support/files/test0.jpg
 - spec/support/files/test1.jpg
 - spec/support/files/test2.jpg