task done

ajit2704/README.md

@@ -0,0 +1,49 @@
+# PROBLEM STATEMENT
+
+### Input 
+
+This image of a graph,having X and Y axes.
+
+![alt tag](input.jpg)
+
+### Desired Output 
+
+Input image + rectangular boxes drawn around the markings on the axes.
+
+![alt tag](output.jpg)
+
+### Languages 
+
+C++ / Python
+
+### Useful tips
+
+Browse through the OpenCV library documentation and find functions that would be helpful to implement this.
+
+The basic idea of solution, is to find contours (regions, maybe `rectangular` ones) that have atleast a certain level of text density and are on the left side of Y-axis or below the X-axis.
+
+Text can be recognised by the sudden change in intensity level with respect to the background.
+
+But the trick lies in differentiating the X & Y axes from the graph's curve - use Google, StackOverflow and OpenCV tutorials to the maximum extent.
+
+It may be tough to get all the readings accurately. but it's fine.
+
+Good luck!
+
+### Important Note
+
+Coordinates of the origin of the graph should not be hard-coded specific to a particular image. It should work on any similar graph image with 2 perpendicular lines.
+
+Variables like cutoff text density can be hard-coded, but it'd be better if the cutoff is dynamically decided.
+
+### Submission Guidelines
+
+(1) Fork this repository on GitHub.
+
+(2) Clone your fork with `git clone` statement.
+
+(3) Make your local changes, in a sub-directory (say, `graph-detect/<username>/`).
+
+(4) Push these changes to your fork with `git add`, `git commit` and `git push` statements.
+
+(5) Create a Pull Request to this repository by comparing across forks with your fork.

ajit2704/auv.py

@@ -0,0 +1,50 @@
+# -*- coding: utf-8 -*-
+"""
+Spyder Editor
+
+This is a temporary script file.
+"""
+
+
+import numpy as np
+import cv2
+
+im = cv2.imread('input.jpg')
+im3 = im.copy()
+
+gray = cv2.cvtColor(im,cv2.COLOR_BGR2GRAY)
+blur = cv2.GaussianBlur(gray,(5,5),0)
+thresh = cv2.adaptiveThreshold(blur,255,1,1,11,2)
+
+
+
+contours,hierarchy = cv2.findContours(thresh.copy(),cv2.RETR_LIST,cv2.CHAIN_APPROX_SIMPLE)
+
+samples =  np.empty((0,100))
+responses = []
+keys = [i for i in range(1,100)]
+
+for cnt in contours:
+    if cv2.contourArea(cnt)>50:
+
+        [x,y,w,h] = cv2.boundingRect(cnt)
+
+        if  h<11.05:
+            cv2.rectangle(im,(x,y),(x+w,y+h),(255,0,0),2)
+            roi = thresh[y:y+h,x:x+w]
+            roismall = cv2.resize(roi,(10,10))
+            cv2.imshow('auv',im)
+            key = cv2.waitKey(0)
+
+
+            if key in keys:
+                responses.append(int(chr(key)))
+                sample = roismall.reshape((1,100))
+                samples = np.append(samples,sample,0)
+
+responses = np.array(responses,np.float32)
+responses = responses.reshape((responses.size,1))
+print ""
+
+np.savetxt('generalsamples.data',samples)
+np.savetxt('generalresponses.data',responses)

auv.py

@@ -0,0 +1,50 @@
+# -*- coding: utf-8 -*-
+"""
+Spyder Editor
+
+This is a temporary script file.
+"""
+
+
+import numpy as np
+import cv2
+
+im = cv2.imread('input.jpg')
+im3 = im.copy()
+
+gray = cv2.cvtColor(im,cv2.COLOR_BGR2GRAY)
+blur = cv2.GaussianBlur(gray,(5,5),0)
+thresh = cv2.adaptiveThreshold(blur,255,1,1,11,2)
+
+
+
+image,contours,hierarchy = cv2.findContours(thresh.copy(),cv2.RETR_LIST,cv2.CHAIN_APPROX_SIMPLE)
+
+samples =  np.empty((0,100))
+responses = []
+keys = [i for i in range(1,100)]
+
+for cnt in contours:
+    if cv2.contourArea(cnt)>50:
+
+        [x,y,w,h] = cv2.boundingRect(cnt)
+
+        if  h<11.05:
+            cv2.rectangle(im,(x,y),(x+w,y+h),(255,0,0),2)
+            roi = thresh[y:y+h,x:x+w]
+            roismall = cv2.resize(roi,(10,10))
+            cv2.imshow('auv',im)
+            key = cv2.waitKey(0)
+
+
+            if key in keys:
+                responses.append(int(chr(key)))
+                sample = roismall.reshape((1,100))
+                samples = np.append(samples,sample,0)
+
+responses = np.array(responses,np.float32)
+responses = responses.reshape((responses.size,1))
+print ""
+
+np.savetxt('generalsamples.data',samples)
+np.savetxt('generalresponses.data',responses)