acfDetect

 Run aggregate channel features object detector on given image(s).

 The input 'I' can either be a single image (or filename) or a cell array
 of images (or filenames). In the first case, the return is a set of bbs
 where each row has the format [x y w h score] and score is the confidence
 of detection. If the input is a cell array, the output is a cell array
 where each element is a set of bbs in the form above (in this case a
 parfor loop is used to speed execution). If 'fileName' is specified, the
 bbs are saved to a comma separated text file and the output is set to
 bbs=1. If saving detections for multiple images the output is stored in
 the format [imgId x y w h score] and imgId is a one-indexed image id.

 A cell of detectors trained with the same channels can be specified,
 detected bbs from each detector are concatenated. If using multiple
 detectors and opts.pNms.separate=1 then each bb has a sixth element
 bbType=j, where j is the j-th detector, see bbNms.m for details.

 USAGE
  bbs = acfDetect( I, detector, [fileName] )

 INPUTS
  I          - input image(s) of filename(s) of input image(s)
  detector   - detector(s) trained via acfTrain
  fileName   - [] target filename (if specified return is 1)

 OUTPUTS
  bbs        - [nx5] array of bounding boxes or cell array of bbs

 EXAMPLE

 See also acfTrain, acfModify, bbGt>loadAll, bbNms

 Piotr's Computer Vision Matlab Toolbox      Version 3.40
 Copyright 2014 Piotr Dollar.  [pdollar-at-gmail.com]
 Licensed under the Simplified BSD License [see external/bsd.txt]