use the detector to find a specific in an image. There should not be space in image names!

##Update Available [logo_name]: 'subway', 'stopsign', 'chase', 'mobil'

• fov_detect added
• Win_stride 1x
• Contrast: 1.2x

##Merge data

• Include 'merge' in your python code.
• Use merge_image(src_dir, res_dir) to do processing.
• Result will be stored in [res_dir]

##Do detection Modify 'code/config.txt' to set the detection threshold and window

• Include 'obj_detect.py', use the 'get_result' function inside.
• Parameters for 'get_result': [img_path], [logo_name], ([store_img=1], [res_dir_path='res/'])
• Return value format:

1. RESULT: empty
2. RESULT: [left_pos]/[right_pos]/[img_cols]/[thres]

• store_img = 1 will store the result in res_dir_path. If you want to make the code run faster, set it to 0

• Use: $ python obj_detect.py [input_dir_path] [logo_name] [threshold] [res_dir_path='res/']
• All the result images will be stored in [res_dir_path], default value is 'res/'
• Recommended threshold between 1.1 and 1.3 (for subway logo)

(in constraction...)

##Detect Field of View This part can detect if the triangulation result is within the field of view from one view points. Support code can be found in 'code/tools/fov_detect'

• Command: $ python fov_detect.py [view_point_file] [triangulation_file] [res_dir]
• Sample of view_point_file and triangulation_file can be found in 'code/sample_data/'

• Result wil be stored in [res_dir]
• Each line in 'result.txt' means [location_number : number_of_points_within_FoV]
• Images are the visualized results.