I guess in image segmentation we are looking for the exact boundaries and it can be considered as pixel-wise or super-pixel-wise classification of points. However, in object detection we are not interested in exact boundaries. Sometime prediction of presence of an object is enough or just a rectangular boundary. In segmentation the outputs are labels for pixels. However, the output in object detection is more categorical.

The two problems are highly correlated, however, there are yet some minor differences in the constraints and objectives. At the end, it will depend on how you formulate the problem.

We can consider salient object detection as a image segmentation because salient image is separated from image background. It is explained with constraints in one of the good paper published in CVPR 2007 -

Learning to detect A Salient Object by Tie Liu, Jian Sun ; Nan-Ning Zheng ; Xiaoou Tang ; Heung-Yeung Shum- ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4270072

I believe image segmentation encompasses a broad area under computer vision. In coherence to the other responses to your question, salient object detection is one of the areas under image segmentation, but mostly two-class segmentation. One of the famous papers on Saliency based segmentation is by Achanta (1). It has been successfully used for video segmentation (2).

In each of the cases, it is mostly pixel/voxel classification i.e. saliency is used as a means to classify a pixel/voxel into foreground or background.

1) R. Achanta, F. Estrada, P. Wils and S. Süsstrunk, "Salient Region Detection and Segmentation", International Conference on Computer Vision Systems (ICVS '08), Vol. 5008, Springer Lecture Notes in Computer Science, pp. 66-75, 2008

2) W. Wang, J. Shen, F. Porikli, "Saliency-aware geodesic video object segmentation," in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp.3395-3402

hi ,

 We can consider saliency detection detection under image segmentation. 

for more info just check the below paper 

W. Wang, J. Shen, F. Porikli, "Saliency-aware geodesic video object segmentation," in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp.3395-3402

Detection and segmentation are two different problems. However, in case of salient object detection process image segmentation can come in handy. Many recent methods have taken advantage of segmentation for predicting salient object region. Consider searching recent CVPR, ICCV, ACM MM, and CGI publication to have a precise idea.

Thanks for all your answers.

if you looking for real-time object detection please read this paper

Conference Paper Automatic, Real Time, Unsupervised Spatio-temporal 3D Object...

Article Weighted Unsupervised Learning for 3D Object Detection